TruongScotl
commited on
Commit
•
0df3adf
1
Parent(s):
d76f8f0
End of training
Browse files- README.md +3 -1
- all_results.json +13 -0
- eval_results.json +7 -0
- train_results.json +9 -0
- trainer_state.json +3437 -0
README.md
CHANGED
@@ -3,6 +3,8 @@ base_model: facebook/musicgen-small
|
|
3 |
library_name: peft
|
4 |
license: cc-by-nc-4.0
|
5 |
tags:
|
|
|
|
|
6 |
- generated_from_trainer
|
7 |
model-index:
|
8 |
- name: musicgen-electronic-lora_airain1
|
@@ -14,7 +16,7 @@ should probably proofread and complete it, then remove this comment. -->
|
|
14 |
|
15 |
# musicgen-electronic-lora_airain1
|
16 |
|
17 |
-
This model is a fine-tuned version of [facebook/musicgen-small](https://huggingface.co/facebook/musicgen-small) on
|
18 |
|
19 |
## Model description
|
20 |
|
|
|
3 |
library_name: peft
|
4 |
license: cc-by-nc-4.0
|
5 |
tags:
|
6 |
+
- text-to-audio
|
7 |
+
- tiny-electronic
|
8 |
- generated_from_trainer
|
9 |
model-index:
|
10 |
- name: musicgen-electronic-lora_airain1
|
|
|
16 |
|
17 |
# musicgen-electronic-lora_airain1
|
18 |
|
19 |
+
This model is a fine-tuned version of [facebook/musicgen-small](https://huggingface.co/facebook/musicgen-small) on the TruongScotl/electronic_music_generating dataset.
|
20 |
|
21 |
## Model description
|
22 |
|
all_results.json
ADDED
@@ -0,0 +1,13 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"epoch": 9.948717948717949,
|
3 |
+
"eval_runtime": 76.8197,
|
4 |
+
"eval_samples": 380,
|
5 |
+
"eval_samples_per_second": 4.947,
|
6 |
+
"eval_steps_per_second": 0.625,
|
7 |
+
"total_flos": 1620553852147860.0,
|
8 |
+
"train_loss": 6.414023149628,
|
9 |
+
"train_runtime": 12031.1209,
|
10 |
+
"train_samples": 1559,
|
11 |
+
"train_samples_per_second": 1.296,
|
12 |
+
"train_steps_per_second": 0.081
|
13 |
+
}
|
eval_results.json
ADDED
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"epoch": 9.948717948717949,
|
3 |
+
"eval_runtime": 76.8197,
|
4 |
+
"eval_samples": 380,
|
5 |
+
"eval_samples_per_second": 4.947,
|
6 |
+
"eval_steps_per_second": 0.625
|
7 |
+
}
|
train_results.json
ADDED
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"epoch": 9.948717948717949,
|
3 |
+
"total_flos": 1620553852147860.0,
|
4 |
+
"train_loss": 6.414023149628,
|
5 |
+
"train_runtime": 12031.1209,
|
6 |
+
"train_samples": 1559,
|
7 |
+
"train_samples_per_second": 1.296,
|
8 |
+
"train_steps_per_second": 0.081
|
9 |
+
}
|
trainer_state.json
ADDED
@@ -0,0 +1,3437 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_metric": null,
|
3 |
+
"best_model_checkpoint": null,
|
4 |
+
"epoch": 9.948717948717949,
|
5 |
+
"eval_steps": 500,
|
6 |
+
"global_step": 970,
|
7 |
+
"is_hyper_param_search": false,
|
8 |
+
"is_local_process_zero": true,
|
9 |
+
"is_world_process_zero": true,
|
10 |
+
"log_history": [
|
11 |
+
{
|
12 |
+
"epoch": 0.020512820512820513,
|
13 |
+
"grad_norm": 7.845447540283203,
|
14 |
+
"learning_rate": 9.989690721649485e-05,
|
15 |
+
"loss": 9.5248,
|
16 |
+
"step": 2
|
17 |
+
},
|
18 |
+
{
|
19 |
+
"epoch": 0.041025641025641026,
|
20 |
+
"grad_norm": 8.389087677001953,
|
21 |
+
"learning_rate": 9.969072164948454e-05,
|
22 |
+
"loss": 9.1949,
|
23 |
+
"step": 4
|
24 |
+
},
|
25 |
+
{
|
26 |
+
"epoch": 0.06153846153846154,
|
27 |
+
"grad_norm": 5.169976711273193,
|
28 |
+
"learning_rate": 9.948453608247423e-05,
|
29 |
+
"loss": 8.6566,
|
30 |
+
"step": 6
|
31 |
+
},
|
32 |
+
{
|
33 |
+
"epoch": 0.08205128205128205,
|
34 |
+
"grad_norm": 4.878049850463867,
|
35 |
+
"learning_rate": 9.927835051546392e-05,
|
36 |
+
"loss": 8.1281,
|
37 |
+
"step": 8
|
38 |
+
},
|
39 |
+
{
|
40 |
+
"epoch": 0.10256410256410256,
|
41 |
+
"grad_norm": 4.062979698181152,
|
42 |
+
"learning_rate": 9.907216494845362e-05,
|
43 |
+
"loss": 7.7464,
|
44 |
+
"step": 10
|
45 |
+
},
|
46 |
+
{
|
47 |
+
"epoch": 0.12307692307692308,
|
48 |
+
"grad_norm": 2.26177716255188,
|
49 |
+
"learning_rate": 9.88659793814433e-05,
|
50 |
+
"loss": 7.4979,
|
51 |
+
"step": 12
|
52 |
+
},
|
53 |
+
{
|
54 |
+
"epoch": 0.14358974358974358,
|
55 |
+
"grad_norm": 1.6796530485153198,
|
56 |
+
"learning_rate": 9.8659793814433e-05,
|
57 |
+
"loss": 7.4101,
|
58 |
+
"step": 14
|
59 |
+
},
|
60 |
+
{
|
61 |
+
"epoch": 0.1641025641025641,
|
62 |
+
"grad_norm": 1.3691627979278564,
|
63 |
+
"learning_rate": 9.845360824742269e-05,
|
64 |
+
"loss": 7.2759,
|
65 |
+
"step": 16
|
66 |
+
},
|
67 |
+
{
|
68 |
+
"epoch": 0.18461538461538463,
|
69 |
+
"grad_norm": 1.8158212900161743,
|
70 |
+
"learning_rate": 9.824742268041237e-05,
|
71 |
+
"loss": 7.1752,
|
72 |
+
"step": 18
|
73 |
+
},
|
74 |
+
{
|
75 |
+
"epoch": 0.20512820512820512,
|
76 |
+
"grad_norm": 1.489174246788025,
|
77 |
+
"learning_rate": 9.804123711340207e-05,
|
78 |
+
"loss": 7.2658,
|
79 |
+
"step": 20
|
80 |
+
},
|
81 |
+
{
|
82 |
+
"epoch": 0.22564102564102564,
|
83 |
+
"grad_norm": 1.363490104675293,
|
84 |
+
"learning_rate": 9.783505154639175e-05,
|
85 |
+
"loss": 7.1867,
|
86 |
+
"step": 22
|
87 |
+
},
|
88 |
+
{
|
89 |
+
"epoch": 0.24615384615384617,
|
90 |
+
"grad_norm": 1.5879547595977783,
|
91 |
+
"learning_rate": 9.762886597938145e-05,
|
92 |
+
"loss": 7.1374,
|
93 |
+
"step": 24
|
94 |
+
},
|
95 |
+
{
|
96 |
+
"epoch": 0.26666666666666666,
|
97 |
+
"grad_norm": 1.046538233757019,
|
98 |
+
"learning_rate": 9.742268041237114e-05,
|
99 |
+
"loss": 7.106,
|
100 |
+
"step": 26
|
101 |
+
},
|
102 |
+
{
|
103 |
+
"epoch": 0.28717948717948716,
|
104 |
+
"grad_norm": 1.169195532798767,
|
105 |
+
"learning_rate": 9.721649484536083e-05,
|
106 |
+
"loss": 7.1945,
|
107 |
+
"step": 28
|
108 |
+
},
|
109 |
+
{
|
110 |
+
"epoch": 0.3076923076923077,
|
111 |
+
"grad_norm": 0.8861628174781799,
|
112 |
+
"learning_rate": 9.701030927835052e-05,
|
113 |
+
"loss": 7.0199,
|
114 |
+
"step": 30
|
115 |
+
},
|
116 |
+
{
|
117 |
+
"epoch": 0.3282051282051282,
|
118 |
+
"grad_norm": 1.4861819744110107,
|
119 |
+
"learning_rate": 9.680412371134021e-05,
|
120 |
+
"loss": 6.8774,
|
121 |
+
"step": 32
|
122 |
+
},
|
123 |
+
{
|
124 |
+
"epoch": 0.3487179487179487,
|
125 |
+
"grad_norm": 0.9394097924232483,
|
126 |
+
"learning_rate": 9.65979381443299e-05,
|
127 |
+
"loss": 6.9136,
|
128 |
+
"step": 34
|
129 |
+
},
|
130 |
+
{
|
131 |
+
"epoch": 0.36923076923076925,
|
132 |
+
"grad_norm": 1.1648708581924438,
|
133 |
+
"learning_rate": 9.639175257731959e-05,
|
134 |
+
"loss": 7.121,
|
135 |
+
"step": 36
|
136 |
+
},
|
137 |
+
{
|
138 |
+
"epoch": 0.38974358974358975,
|
139 |
+
"grad_norm": 0.8142098784446716,
|
140 |
+
"learning_rate": 9.618556701030928e-05,
|
141 |
+
"loss": 7.1084,
|
142 |
+
"step": 38
|
143 |
+
},
|
144 |
+
{
|
145 |
+
"epoch": 0.41025641025641024,
|
146 |
+
"grad_norm": 1.0768697261810303,
|
147 |
+
"learning_rate": 9.597938144329898e-05,
|
148 |
+
"loss": 7.0565,
|
149 |
+
"step": 40
|
150 |
+
},
|
151 |
+
{
|
152 |
+
"epoch": 0.4307692307692308,
|
153 |
+
"grad_norm": 0.6553409099578857,
|
154 |
+
"learning_rate": 9.577319587628867e-05,
|
155 |
+
"loss": 6.9963,
|
156 |
+
"step": 42
|
157 |
+
},
|
158 |
+
{
|
159 |
+
"epoch": 0.4512820512820513,
|
160 |
+
"grad_norm": 0.8519483208656311,
|
161 |
+
"learning_rate": 9.556701030927836e-05,
|
162 |
+
"loss": 7.0575,
|
163 |
+
"step": 44
|
164 |
+
},
|
165 |
+
{
|
166 |
+
"epoch": 0.4717948717948718,
|
167 |
+
"grad_norm": 1.1110036373138428,
|
168 |
+
"learning_rate": 9.536082474226805e-05,
|
169 |
+
"loss": 6.998,
|
170 |
+
"step": 46
|
171 |
+
},
|
172 |
+
{
|
173 |
+
"epoch": 0.49230769230769234,
|
174 |
+
"grad_norm": 0.8143446445465088,
|
175 |
+
"learning_rate": 9.515463917525773e-05,
|
176 |
+
"loss": 6.9762,
|
177 |
+
"step": 48
|
178 |
+
},
|
179 |
+
{
|
180 |
+
"epoch": 0.5128205128205128,
|
181 |
+
"grad_norm": 1.4277548789978027,
|
182 |
+
"learning_rate": 9.494845360824743e-05,
|
183 |
+
"loss": 6.9832,
|
184 |
+
"step": 50
|
185 |
+
},
|
186 |
+
{
|
187 |
+
"epoch": 0.5333333333333333,
|
188 |
+
"grad_norm": 0.6880565881729126,
|
189 |
+
"learning_rate": 9.474226804123711e-05,
|
190 |
+
"loss": 7.0254,
|
191 |
+
"step": 52
|
192 |
+
},
|
193 |
+
{
|
194 |
+
"epoch": 0.5538461538461539,
|
195 |
+
"grad_norm": 0.8316001296043396,
|
196 |
+
"learning_rate": 9.45360824742268e-05,
|
197 |
+
"loss": 6.9704,
|
198 |
+
"step": 54
|
199 |
+
},
|
200 |
+
{
|
201 |
+
"epoch": 0.5743589743589743,
|
202 |
+
"grad_norm": 0.91054368019104,
|
203 |
+
"learning_rate": 9.43298969072165e-05,
|
204 |
+
"loss": 6.9184,
|
205 |
+
"step": 56
|
206 |
+
},
|
207 |
+
{
|
208 |
+
"epoch": 0.5948717948717949,
|
209 |
+
"grad_norm": 0.7943957448005676,
|
210 |
+
"learning_rate": 9.412371134020619e-05,
|
211 |
+
"loss": 6.9518,
|
212 |
+
"step": 58
|
213 |
+
},
|
214 |
+
{
|
215 |
+
"epoch": 0.6153846153846154,
|
216 |
+
"grad_norm": 1.0420604944229126,
|
217 |
+
"learning_rate": 9.391752577319588e-05,
|
218 |
+
"loss": 6.9369,
|
219 |
+
"step": 60
|
220 |
+
},
|
221 |
+
{
|
222 |
+
"epoch": 0.6358974358974359,
|
223 |
+
"grad_norm": 0.7538536787033081,
|
224 |
+
"learning_rate": 9.371134020618557e-05,
|
225 |
+
"loss": 6.8748,
|
226 |
+
"step": 62
|
227 |
+
},
|
228 |
+
{
|
229 |
+
"epoch": 0.6564102564102564,
|
230 |
+
"grad_norm": 1.2202829122543335,
|
231 |
+
"learning_rate": 9.350515463917526e-05,
|
232 |
+
"loss": 6.8046,
|
233 |
+
"step": 64
|
234 |
+
},
|
235 |
+
{
|
236 |
+
"epoch": 0.676923076923077,
|
237 |
+
"grad_norm": 1.0980966091156006,
|
238 |
+
"learning_rate": 9.329896907216495e-05,
|
239 |
+
"loss": 6.8146,
|
240 |
+
"step": 66
|
241 |
+
},
|
242 |
+
{
|
243 |
+
"epoch": 0.6974358974358974,
|
244 |
+
"grad_norm": 1.5506396293640137,
|
245 |
+
"learning_rate": 9.309278350515465e-05,
|
246 |
+
"loss": 7.0011,
|
247 |
+
"step": 68
|
248 |
+
},
|
249 |
+
{
|
250 |
+
"epoch": 0.717948717948718,
|
251 |
+
"grad_norm": 1.1768453121185303,
|
252 |
+
"learning_rate": 9.288659793814434e-05,
|
253 |
+
"loss": 6.8345,
|
254 |
+
"step": 70
|
255 |
+
},
|
256 |
+
{
|
257 |
+
"epoch": 0.7384615384615385,
|
258 |
+
"grad_norm": 1.1223098039627075,
|
259 |
+
"learning_rate": 9.268041237113403e-05,
|
260 |
+
"loss": 6.8371,
|
261 |
+
"step": 72
|
262 |
+
},
|
263 |
+
{
|
264 |
+
"epoch": 0.7589743589743589,
|
265 |
+
"grad_norm": 0.7969951629638672,
|
266 |
+
"learning_rate": 9.247422680412372e-05,
|
267 |
+
"loss": 6.902,
|
268 |
+
"step": 74
|
269 |
+
},
|
270 |
+
{
|
271 |
+
"epoch": 0.7794871794871795,
|
272 |
+
"grad_norm": 0.8236146569252014,
|
273 |
+
"learning_rate": 9.22680412371134e-05,
|
274 |
+
"loss": 6.8724,
|
275 |
+
"step": 76
|
276 |
+
},
|
277 |
+
{
|
278 |
+
"epoch": 0.8,
|
279 |
+
"grad_norm": 1.0809398889541626,
|
280 |
+
"learning_rate": 9.206185567010309e-05,
|
281 |
+
"loss": 6.8455,
|
282 |
+
"step": 78
|
283 |
+
},
|
284 |
+
{
|
285 |
+
"epoch": 0.8205128205128205,
|
286 |
+
"grad_norm": 1.0138286352157593,
|
287 |
+
"learning_rate": 9.185567010309279e-05,
|
288 |
+
"loss": 6.8771,
|
289 |
+
"step": 80
|
290 |
+
},
|
291 |
+
{
|
292 |
+
"epoch": 0.841025641025641,
|
293 |
+
"grad_norm": 1.5918043851852417,
|
294 |
+
"learning_rate": 9.164948453608247e-05,
|
295 |
+
"loss": 6.8593,
|
296 |
+
"step": 82
|
297 |
+
},
|
298 |
+
{
|
299 |
+
"epoch": 0.8615384615384616,
|
300 |
+
"grad_norm": 1.9432188272476196,
|
301 |
+
"learning_rate": 9.144329896907217e-05,
|
302 |
+
"loss": 6.8676,
|
303 |
+
"step": 84
|
304 |
+
},
|
305 |
+
{
|
306 |
+
"epoch": 0.882051282051282,
|
307 |
+
"grad_norm": 1.5236680507659912,
|
308 |
+
"learning_rate": 9.123711340206186e-05,
|
309 |
+
"loss": 6.6954,
|
310 |
+
"step": 86
|
311 |
+
},
|
312 |
+
{
|
313 |
+
"epoch": 0.9025641025641026,
|
314 |
+
"grad_norm": 1.347008228302002,
|
315 |
+
"learning_rate": 9.103092783505155e-05,
|
316 |
+
"loss": 6.7739,
|
317 |
+
"step": 88
|
318 |
+
},
|
319 |
+
{
|
320 |
+
"epoch": 0.9230769230769231,
|
321 |
+
"grad_norm": 0.8976010680198669,
|
322 |
+
"learning_rate": 9.082474226804124e-05,
|
323 |
+
"loss": 6.8272,
|
324 |
+
"step": 90
|
325 |
+
},
|
326 |
+
{
|
327 |
+
"epoch": 0.9435897435897436,
|
328 |
+
"grad_norm": 1.0713074207305908,
|
329 |
+
"learning_rate": 9.061855670103093e-05,
|
330 |
+
"loss": 6.722,
|
331 |
+
"step": 92
|
332 |
+
},
|
333 |
+
{
|
334 |
+
"epoch": 0.9641025641025641,
|
335 |
+
"grad_norm": 1.3298990726470947,
|
336 |
+
"learning_rate": 9.041237113402063e-05,
|
337 |
+
"loss": 6.7779,
|
338 |
+
"step": 94
|
339 |
+
},
|
340 |
+
{
|
341 |
+
"epoch": 0.9846153846153847,
|
342 |
+
"grad_norm": 1.3666430711746216,
|
343 |
+
"learning_rate": 9.020618556701031e-05,
|
344 |
+
"loss": 6.6804,
|
345 |
+
"step": 96
|
346 |
+
},
|
347 |
+
{
|
348 |
+
"epoch": 1.005128205128205,
|
349 |
+
"grad_norm": 1.5196622610092163,
|
350 |
+
"learning_rate": 9e-05,
|
351 |
+
"loss": 6.7935,
|
352 |
+
"step": 98
|
353 |
+
},
|
354 |
+
{
|
355 |
+
"epoch": 1.0256410256410255,
|
356 |
+
"grad_norm": 1.0106366872787476,
|
357 |
+
"learning_rate": 8.97938144329897e-05,
|
358 |
+
"loss": 6.7578,
|
359 |
+
"step": 100
|
360 |
+
},
|
361 |
+
{
|
362 |
+
"epoch": 1.0461538461538462,
|
363 |
+
"grad_norm": 0.9565912485122681,
|
364 |
+
"learning_rate": 8.958762886597939e-05,
|
365 |
+
"loss": 6.81,
|
366 |
+
"step": 102
|
367 |
+
},
|
368 |
+
{
|
369 |
+
"epoch": 1.0666666666666667,
|
370 |
+
"grad_norm": 0.7427358627319336,
|
371 |
+
"learning_rate": 8.938144329896908e-05,
|
372 |
+
"loss": 6.7069,
|
373 |
+
"step": 104
|
374 |
+
},
|
375 |
+
{
|
376 |
+
"epoch": 1.087179487179487,
|
377 |
+
"grad_norm": 1.8422271013259888,
|
378 |
+
"learning_rate": 8.917525773195877e-05,
|
379 |
+
"loss": 6.7098,
|
380 |
+
"step": 106
|
381 |
+
},
|
382 |
+
{
|
383 |
+
"epoch": 1.1076923076923078,
|
384 |
+
"grad_norm": 0.5381059050559998,
|
385 |
+
"learning_rate": 8.896907216494845e-05,
|
386 |
+
"loss": 6.7934,
|
387 |
+
"step": 108
|
388 |
+
},
|
389 |
+
{
|
390 |
+
"epoch": 1.1282051282051282,
|
391 |
+
"grad_norm": 0.8050350546836853,
|
392 |
+
"learning_rate": 8.876288659793815e-05,
|
393 |
+
"loss": 6.6336,
|
394 |
+
"step": 110
|
395 |
+
},
|
396 |
+
{
|
397 |
+
"epoch": 1.1487179487179486,
|
398 |
+
"grad_norm": 0.570678174495697,
|
399 |
+
"learning_rate": 8.855670103092783e-05,
|
400 |
+
"loss": 6.8265,
|
401 |
+
"step": 112
|
402 |
+
},
|
403 |
+
{
|
404 |
+
"epoch": 1.1692307692307693,
|
405 |
+
"grad_norm": 0.5224724411964417,
|
406 |
+
"learning_rate": 8.835051546391753e-05,
|
407 |
+
"loss": 6.9796,
|
408 |
+
"step": 114
|
409 |
+
},
|
410 |
+
{
|
411 |
+
"epoch": 1.1897435897435897,
|
412 |
+
"grad_norm": 0.6684609651565552,
|
413 |
+
"learning_rate": 8.814432989690722e-05,
|
414 |
+
"loss": 6.701,
|
415 |
+
"step": 116
|
416 |
+
},
|
417 |
+
{
|
418 |
+
"epoch": 1.2102564102564102,
|
419 |
+
"grad_norm": 0.9216952919960022,
|
420 |
+
"learning_rate": 8.793814432989691e-05,
|
421 |
+
"loss": 6.819,
|
422 |
+
"step": 118
|
423 |
+
},
|
424 |
+
{
|
425 |
+
"epoch": 1.2307692307692308,
|
426 |
+
"grad_norm": 0.925770103931427,
|
427 |
+
"learning_rate": 8.77319587628866e-05,
|
428 |
+
"loss": 6.6087,
|
429 |
+
"step": 120
|
430 |
+
},
|
431 |
+
{
|
432 |
+
"epoch": 1.2512820512820513,
|
433 |
+
"grad_norm": 0.6525481343269348,
|
434 |
+
"learning_rate": 8.752577319587629e-05,
|
435 |
+
"loss": 6.7305,
|
436 |
+
"step": 122
|
437 |
+
},
|
438 |
+
{
|
439 |
+
"epoch": 1.2717948717948717,
|
440 |
+
"grad_norm": 0.7587235569953918,
|
441 |
+
"learning_rate": 8.731958762886599e-05,
|
442 |
+
"loss": 6.6621,
|
443 |
+
"step": 124
|
444 |
+
},
|
445 |
+
{
|
446 |
+
"epoch": 1.2923076923076924,
|
447 |
+
"grad_norm": 0.7010625600814819,
|
448 |
+
"learning_rate": 8.711340206185567e-05,
|
449 |
+
"loss": 6.6443,
|
450 |
+
"step": 126
|
451 |
+
},
|
452 |
+
{
|
453 |
+
"epoch": 1.3128205128205128,
|
454 |
+
"grad_norm": 0.7843243479728699,
|
455 |
+
"learning_rate": 8.690721649484537e-05,
|
456 |
+
"loss": 6.7056,
|
457 |
+
"step": 128
|
458 |
+
},
|
459 |
+
{
|
460 |
+
"epoch": 1.3333333333333333,
|
461 |
+
"grad_norm": 0.8412699699401855,
|
462 |
+
"learning_rate": 8.670103092783506e-05,
|
463 |
+
"loss": 6.8242,
|
464 |
+
"step": 130
|
465 |
+
},
|
466 |
+
{
|
467 |
+
"epoch": 1.353846153846154,
|
468 |
+
"grad_norm": 0.899800181388855,
|
469 |
+
"learning_rate": 8.649484536082475e-05,
|
470 |
+
"loss": 6.7249,
|
471 |
+
"step": 132
|
472 |
+
},
|
473 |
+
{
|
474 |
+
"epoch": 1.3743589743589744,
|
475 |
+
"grad_norm": 1.413313388824463,
|
476 |
+
"learning_rate": 8.628865979381444e-05,
|
477 |
+
"loss": 6.6534,
|
478 |
+
"step": 134
|
479 |
+
},
|
480 |
+
{
|
481 |
+
"epoch": 1.3948717948717948,
|
482 |
+
"grad_norm": 1.53885018825531,
|
483 |
+
"learning_rate": 8.608247422680413e-05,
|
484 |
+
"loss": 6.6828,
|
485 |
+
"step": 136
|
486 |
+
},
|
487 |
+
{
|
488 |
+
"epoch": 1.4153846153846155,
|
489 |
+
"grad_norm": 0.7499423027038574,
|
490 |
+
"learning_rate": 8.587628865979381e-05,
|
491 |
+
"loss": 6.5888,
|
492 |
+
"step": 138
|
493 |
+
},
|
494 |
+
{
|
495 |
+
"epoch": 1.435897435897436,
|
496 |
+
"grad_norm": 0.48903796076774597,
|
497 |
+
"learning_rate": 8.567010309278351e-05,
|
498 |
+
"loss": 6.7217,
|
499 |
+
"step": 140
|
500 |
+
},
|
501 |
+
{
|
502 |
+
"epoch": 1.4564102564102563,
|
503 |
+
"grad_norm": 0.5164963006973267,
|
504 |
+
"learning_rate": 8.546391752577319e-05,
|
505 |
+
"loss": 6.7929,
|
506 |
+
"step": 142
|
507 |
+
},
|
508 |
+
{
|
509 |
+
"epoch": 1.476923076923077,
|
510 |
+
"grad_norm": 0.4488673210144043,
|
511 |
+
"learning_rate": 8.525773195876289e-05,
|
512 |
+
"loss": 6.7542,
|
513 |
+
"step": 144
|
514 |
+
},
|
515 |
+
{
|
516 |
+
"epoch": 1.4974358974358974,
|
517 |
+
"grad_norm": 0.9296787977218628,
|
518 |
+
"learning_rate": 8.505154639175259e-05,
|
519 |
+
"loss": 6.595,
|
520 |
+
"step": 146
|
521 |
+
},
|
522 |
+
{
|
523 |
+
"epoch": 1.5179487179487179,
|
524 |
+
"grad_norm": 0.48948934674263,
|
525 |
+
"learning_rate": 8.484536082474227e-05,
|
526 |
+
"loss": 6.6336,
|
527 |
+
"step": 148
|
528 |
+
},
|
529 |
+
{
|
530 |
+
"epoch": 1.5384615384615383,
|
531 |
+
"grad_norm": 1.3344464302062988,
|
532 |
+
"learning_rate": 8.463917525773197e-05,
|
533 |
+
"loss": 6.5617,
|
534 |
+
"step": 150
|
535 |
+
},
|
536 |
+
{
|
537 |
+
"epoch": 1.558974358974359,
|
538 |
+
"grad_norm": 0.8237192034721375,
|
539 |
+
"learning_rate": 8.443298969072165e-05,
|
540 |
+
"loss": 6.6697,
|
541 |
+
"step": 152
|
542 |
+
},
|
543 |
+
{
|
544 |
+
"epoch": 1.5794871794871796,
|
545 |
+
"grad_norm": 1.013886570930481,
|
546 |
+
"learning_rate": 8.422680412371135e-05,
|
547 |
+
"loss": 6.5097,
|
548 |
+
"step": 154
|
549 |
+
},
|
550 |
+
{
|
551 |
+
"epoch": 1.6,
|
552 |
+
"grad_norm": 0.5843191742897034,
|
553 |
+
"learning_rate": 8.402061855670103e-05,
|
554 |
+
"loss": 6.761,
|
555 |
+
"step": 156
|
556 |
+
},
|
557 |
+
{
|
558 |
+
"epoch": 1.6205128205128205,
|
559 |
+
"grad_norm": 0.5963282585144043,
|
560 |
+
"learning_rate": 8.381443298969073e-05,
|
561 |
+
"loss": 6.8233,
|
562 |
+
"step": 158
|
563 |
+
},
|
564 |
+
{
|
565 |
+
"epoch": 1.641025641025641,
|
566 |
+
"grad_norm": 0.7131335735321045,
|
567 |
+
"learning_rate": 8.360824742268042e-05,
|
568 |
+
"loss": 6.5581,
|
569 |
+
"step": 160
|
570 |
+
},
|
571 |
+
{
|
572 |
+
"epoch": 1.6615384615384614,
|
573 |
+
"grad_norm": 0.5245983004570007,
|
574 |
+
"learning_rate": 8.340206185567011e-05,
|
575 |
+
"loss": 6.5801,
|
576 |
+
"step": 162
|
577 |
+
},
|
578 |
+
{
|
579 |
+
"epoch": 1.682051282051282,
|
580 |
+
"grad_norm": 0.6111565232276917,
|
581 |
+
"learning_rate": 8.31958762886598e-05,
|
582 |
+
"loss": 6.6157,
|
583 |
+
"step": 164
|
584 |
+
},
|
585 |
+
{
|
586 |
+
"epoch": 1.7025641025641025,
|
587 |
+
"grad_norm": 0.7455165982246399,
|
588 |
+
"learning_rate": 8.298969072164949e-05,
|
589 |
+
"loss": 6.6316,
|
590 |
+
"step": 166
|
591 |
+
},
|
592 |
+
{
|
593 |
+
"epoch": 1.7230769230769232,
|
594 |
+
"grad_norm": 0.7735164761543274,
|
595 |
+
"learning_rate": 8.278350515463917e-05,
|
596 |
+
"loss": 6.6299,
|
597 |
+
"step": 168
|
598 |
+
},
|
599 |
+
{
|
600 |
+
"epoch": 1.7435897435897436,
|
601 |
+
"grad_norm": 0.5676758885383606,
|
602 |
+
"learning_rate": 8.257731958762887e-05,
|
603 |
+
"loss": 6.7204,
|
604 |
+
"step": 170
|
605 |
+
},
|
606 |
+
{
|
607 |
+
"epoch": 1.764102564102564,
|
608 |
+
"grad_norm": 0.8821406364440918,
|
609 |
+
"learning_rate": 8.237113402061855e-05,
|
610 |
+
"loss": 6.3924,
|
611 |
+
"step": 172
|
612 |
+
},
|
613 |
+
{
|
614 |
+
"epoch": 1.7846153846153845,
|
615 |
+
"grad_norm": 0.6726123094558716,
|
616 |
+
"learning_rate": 8.216494845360825e-05,
|
617 |
+
"loss": 6.7632,
|
618 |
+
"step": 174
|
619 |
+
},
|
620 |
+
{
|
621 |
+
"epoch": 1.8051282051282052,
|
622 |
+
"grad_norm": 0.5056736469268799,
|
623 |
+
"learning_rate": 8.195876288659795e-05,
|
624 |
+
"loss": 6.5864,
|
625 |
+
"step": 176
|
626 |
+
},
|
627 |
+
{
|
628 |
+
"epoch": 1.8256410256410256,
|
629 |
+
"grad_norm": 0.5093751549720764,
|
630 |
+
"learning_rate": 8.175257731958763e-05,
|
631 |
+
"loss": 6.5112,
|
632 |
+
"step": 178
|
633 |
+
},
|
634 |
+
{
|
635 |
+
"epoch": 1.8461538461538463,
|
636 |
+
"grad_norm": 0.6652563810348511,
|
637 |
+
"learning_rate": 8.154639175257733e-05,
|
638 |
+
"loss": 6.5538,
|
639 |
+
"step": 180
|
640 |
+
},
|
641 |
+
{
|
642 |
+
"epoch": 1.8666666666666667,
|
643 |
+
"grad_norm": 0.6594399213790894,
|
644 |
+
"learning_rate": 8.134020618556701e-05,
|
645 |
+
"loss": 6.6136,
|
646 |
+
"step": 182
|
647 |
+
},
|
648 |
+
{
|
649 |
+
"epoch": 1.8871794871794871,
|
650 |
+
"grad_norm": 0.6552271246910095,
|
651 |
+
"learning_rate": 8.113402061855671e-05,
|
652 |
+
"loss": 6.4301,
|
653 |
+
"step": 184
|
654 |
+
},
|
655 |
+
{
|
656 |
+
"epoch": 1.9076923076923076,
|
657 |
+
"grad_norm": 1.0746954679489136,
|
658 |
+
"learning_rate": 8.092783505154639e-05,
|
659 |
+
"loss": 6.3992,
|
660 |
+
"step": 186
|
661 |
+
},
|
662 |
+
{
|
663 |
+
"epoch": 1.9282051282051282,
|
664 |
+
"grad_norm": 0.5134009122848511,
|
665 |
+
"learning_rate": 8.072164948453609e-05,
|
666 |
+
"loss": 6.5104,
|
667 |
+
"step": 188
|
668 |
+
},
|
669 |
+
{
|
670 |
+
"epoch": 1.9487179487179487,
|
671 |
+
"grad_norm": 0.6844182014465332,
|
672 |
+
"learning_rate": 8.051546391752578e-05,
|
673 |
+
"loss": 6.4957,
|
674 |
+
"step": 190
|
675 |
+
},
|
676 |
+
{
|
677 |
+
"epoch": 1.9692307692307693,
|
678 |
+
"grad_norm": 0.5666801929473877,
|
679 |
+
"learning_rate": 8.030927835051547e-05,
|
680 |
+
"loss": 6.679,
|
681 |
+
"step": 192
|
682 |
+
},
|
683 |
+
{
|
684 |
+
"epoch": 1.9897435897435898,
|
685 |
+
"grad_norm": 0.7489727139472961,
|
686 |
+
"learning_rate": 8.010309278350516e-05,
|
687 |
+
"loss": 6.5199,
|
688 |
+
"step": 194
|
689 |
+
},
|
690 |
+
{
|
691 |
+
"epoch": 2.01025641025641,
|
692 |
+
"grad_norm": 0.6453472971916199,
|
693 |
+
"learning_rate": 7.989690721649485e-05,
|
694 |
+
"loss": 6.5507,
|
695 |
+
"step": 196
|
696 |
+
},
|
697 |
+
{
|
698 |
+
"epoch": 2.0307692307692307,
|
699 |
+
"grad_norm": 0.5807784199714661,
|
700 |
+
"learning_rate": 7.969072164948455e-05,
|
701 |
+
"loss": 6.7443,
|
702 |
+
"step": 198
|
703 |
+
},
|
704 |
+
{
|
705 |
+
"epoch": 2.051282051282051,
|
706 |
+
"grad_norm": 0.6182963848114014,
|
707 |
+
"learning_rate": 7.948453608247423e-05,
|
708 |
+
"loss": 6.4959,
|
709 |
+
"step": 200
|
710 |
+
},
|
711 |
+
{
|
712 |
+
"epoch": 2.071794871794872,
|
713 |
+
"grad_norm": 0.579248309135437,
|
714 |
+
"learning_rate": 7.927835051546391e-05,
|
715 |
+
"loss": 6.597,
|
716 |
+
"step": 202
|
717 |
+
},
|
718 |
+
{
|
719 |
+
"epoch": 2.0923076923076924,
|
720 |
+
"grad_norm": 0.5996450781822205,
|
721 |
+
"learning_rate": 7.907216494845361e-05,
|
722 |
+
"loss": 6.7087,
|
723 |
+
"step": 204
|
724 |
+
},
|
725 |
+
{
|
726 |
+
"epoch": 2.112820512820513,
|
727 |
+
"grad_norm": 0.4910239577293396,
|
728 |
+
"learning_rate": 7.88659793814433e-05,
|
729 |
+
"loss": 6.666,
|
730 |
+
"step": 206
|
731 |
+
},
|
732 |
+
{
|
733 |
+
"epoch": 2.1333333333333333,
|
734 |
+
"grad_norm": 0.6239397525787354,
|
735 |
+
"learning_rate": 7.865979381443299e-05,
|
736 |
+
"loss": 6.4757,
|
737 |
+
"step": 208
|
738 |
+
},
|
739 |
+
{
|
740 |
+
"epoch": 2.1538461538461537,
|
741 |
+
"grad_norm": 0.581991970539093,
|
742 |
+
"learning_rate": 7.845360824742269e-05,
|
743 |
+
"loss": 6.4388,
|
744 |
+
"step": 210
|
745 |
+
},
|
746 |
+
{
|
747 |
+
"epoch": 2.174358974358974,
|
748 |
+
"grad_norm": 0.6894115209579468,
|
749 |
+
"learning_rate": 7.824742268041237e-05,
|
750 |
+
"loss": 6.7599,
|
751 |
+
"step": 212
|
752 |
+
},
|
753 |
+
{
|
754 |
+
"epoch": 2.194871794871795,
|
755 |
+
"grad_norm": 0.6936022043228149,
|
756 |
+
"learning_rate": 7.804123711340207e-05,
|
757 |
+
"loss": 6.5268,
|
758 |
+
"step": 214
|
759 |
+
},
|
760 |
+
{
|
761 |
+
"epoch": 2.2153846153846155,
|
762 |
+
"grad_norm": 0.6604471802711487,
|
763 |
+
"learning_rate": 7.783505154639175e-05,
|
764 |
+
"loss": 6.5513,
|
765 |
+
"step": 216
|
766 |
+
},
|
767 |
+
{
|
768 |
+
"epoch": 2.235897435897436,
|
769 |
+
"grad_norm": 0.6073073744773865,
|
770 |
+
"learning_rate": 7.762886597938145e-05,
|
771 |
+
"loss": 6.5175,
|
772 |
+
"step": 218
|
773 |
+
},
|
774 |
+
{
|
775 |
+
"epoch": 2.2564102564102564,
|
776 |
+
"grad_norm": 0.5706610679626465,
|
777 |
+
"learning_rate": 7.742268041237114e-05,
|
778 |
+
"loss": 6.6232,
|
779 |
+
"step": 220
|
780 |
+
},
|
781 |
+
{
|
782 |
+
"epoch": 2.276923076923077,
|
783 |
+
"grad_norm": 0.6433441042900085,
|
784 |
+
"learning_rate": 7.721649484536083e-05,
|
785 |
+
"loss": 6.324,
|
786 |
+
"step": 222
|
787 |
+
},
|
788 |
+
{
|
789 |
+
"epoch": 2.2974358974358973,
|
790 |
+
"grad_norm": 0.6643648743629456,
|
791 |
+
"learning_rate": 7.701030927835053e-05,
|
792 |
+
"loss": 6.5876,
|
793 |
+
"step": 224
|
794 |
+
},
|
795 |
+
{
|
796 |
+
"epoch": 2.3179487179487177,
|
797 |
+
"grad_norm": 0.6086193323135376,
|
798 |
+
"learning_rate": 7.680412371134021e-05,
|
799 |
+
"loss": 6.5513,
|
800 |
+
"step": 226
|
801 |
+
},
|
802 |
+
{
|
803 |
+
"epoch": 2.3384615384615386,
|
804 |
+
"grad_norm": 0.7945834994316101,
|
805 |
+
"learning_rate": 7.65979381443299e-05,
|
806 |
+
"loss": 6.4032,
|
807 |
+
"step": 228
|
808 |
+
},
|
809 |
+
{
|
810 |
+
"epoch": 2.358974358974359,
|
811 |
+
"grad_norm": 0.7115459442138672,
|
812 |
+
"learning_rate": 7.639175257731959e-05,
|
813 |
+
"loss": 6.7107,
|
814 |
+
"step": 230
|
815 |
+
},
|
816 |
+
{
|
817 |
+
"epoch": 2.3794871794871795,
|
818 |
+
"grad_norm": 0.6793861985206604,
|
819 |
+
"learning_rate": 7.618556701030927e-05,
|
820 |
+
"loss": 6.6648,
|
821 |
+
"step": 232
|
822 |
+
},
|
823 |
+
{
|
824 |
+
"epoch": 2.4,
|
825 |
+
"grad_norm": 0.7627358436584473,
|
826 |
+
"learning_rate": 7.597938144329897e-05,
|
827 |
+
"loss": 6.5648,
|
828 |
+
"step": 234
|
829 |
+
},
|
830 |
+
{
|
831 |
+
"epoch": 2.4205128205128204,
|
832 |
+
"grad_norm": 0.6426063179969788,
|
833 |
+
"learning_rate": 7.577319587628867e-05,
|
834 |
+
"loss": 6.2766,
|
835 |
+
"step": 236
|
836 |
+
},
|
837 |
+
{
|
838 |
+
"epoch": 2.4410256410256412,
|
839 |
+
"grad_norm": 0.6580905914306641,
|
840 |
+
"learning_rate": 7.556701030927835e-05,
|
841 |
+
"loss": 6.6025,
|
842 |
+
"step": 238
|
843 |
+
},
|
844 |
+
{
|
845 |
+
"epoch": 2.4615384615384617,
|
846 |
+
"grad_norm": 0.7617653608322144,
|
847 |
+
"learning_rate": 7.536082474226805e-05,
|
848 |
+
"loss": 6.5868,
|
849 |
+
"step": 240
|
850 |
+
},
|
851 |
+
{
|
852 |
+
"epoch": 2.482051282051282,
|
853 |
+
"grad_norm": 0.5646825432777405,
|
854 |
+
"learning_rate": 7.515463917525773e-05,
|
855 |
+
"loss": 6.6345,
|
856 |
+
"step": 242
|
857 |
+
},
|
858 |
+
{
|
859 |
+
"epoch": 2.5025641025641026,
|
860 |
+
"grad_norm": 0.637170135974884,
|
861 |
+
"learning_rate": 7.494845360824743e-05,
|
862 |
+
"loss": 6.4776,
|
863 |
+
"step": 244
|
864 |
+
},
|
865 |
+
{
|
866 |
+
"epoch": 2.523076923076923,
|
867 |
+
"grad_norm": 0.6873669028282166,
|
868 |
+
"learning_rate": 7.474226804123711e-05,
|
869 |
+
"loss": 6.5078,
|
870 |
+
"step": 246
|
871 |
+
},
|
872 |
+
{
|
873 |
+
"epoch": 2.5435897435897434,
|
874 |
+
"grad_norm": 0.5859996676445007,
|
875 |
+
"learning_rate": 7.453608247422681e-05,
|
876 |
+
"loss": 6.3846,
|
877 |
+
"step": 248
|
878 |
+
},
|
879 |
+
{
|
880 |
+
"epoch": 2.564102564102564,
|
881 |
+
"grad_norm": 0.760138750076294,
|
882 |
+
"learning_rate": 7.43298969072165e-05,
|
883 |
+
"loss": 6.4852,
|
884 |
+
"step": 250
|
885 |
+
},
|
886 |
+
{
|
887 |
+
"epoch": 2.5846153846153848,
|
888 |
+
"grad_norm": 0.650406002998352,
|
889 |
+
"learning_rate": 7.412371134020619e-05,
|
890 |
+
"loss": 6.57,
|
891 |
+
"step": 252
|
892 |
+
},
|
893 |
+
{
|
894 |
+
"epoch": 2.605128205128205,
|
895 |
+
"grad_norm": 0.591342031955719,
|
896 |
+
"learning_rate": 7.391752577319589e-05,
|
897 |
+
"loss": 6.5708,
|
898 |
+
"step": 254
|
899 |
+
},
|
900 |
+
{
|
901 |
+
"epoch": 2.6256410256410256,
|
902 |
+
"grad_norm": 0.5927435755729675,
|
903 |
+
"learning_rate": 7.371134020618557e-05,
|
904 |
+
"loss": 6.6123,
|
905 |
+
"step": 256
|
906 |
+
},
|
907 |
+
{
|
908 |
+
"epoch": 2.646153846153846,
|
909 |
+
"grad_norm": 0.7065881490707397,
|
910 |
+
"learning_rate": 7.350515463917527e-05,
|
911 |
+
"loss": 6.5514,
|
912 |
+
"step": 258
|
913 |
+
},
|
914 |
+
{
|
915 |
+
"epoch": 2.6666666666666665,
|
916 |
+
"grad_norm": 0.4774336516857147,
|
917 |
+
"learning_rate": 7.329896907216495e-05,
|
918 |
+
"loss": 6.4391,
|
919 |
+
"step": 260
|
920 |
+
},
|
921 |
+
{
|
922 |
+
"epoch": 2.6871794871794874,
|
923 |
+
"grad_norm": 0.6040366888046265,
|
924 |
+
"learning_rate": 7.309278350515463e-05,
|
925 |
+
"loss": 6.5542,
|
926 |
+
"step": 262
|
927 |
+
},
|
928 |
+
{
|
929 |
+
"epoch": 2.707692307692308,
|
930 |
+
"grad_norm": 0.6922281980514526,
|
931 |
+
"learning_rate": 7.288659793814433e-05,
|
932 |
+
"loss": 6.4446,
|
933 |
+
"step": 264
|
934 |
+
},
|
935 |
+
{
|
936 |
+
"epoch": 2.7282051282051283,
|
937 |
+
"grad_norm": 0.8469168543815613,
|
938 |
+
"learning_rate": 7.268041237113403e-05,
|
939 |
+
"loss": 6.4419,
|
940 |
+
"step": 266
|
941 |
+
},
|
942 |
+
{
|
943 |
+
"epoch": 2.7487179487179487,
|
944 |
+
"grad_norm": 0.6104958057403564,
|
945 |
+
"learning_rate": 7.247422680412371e-05,
|
946 |
+
"loss": 6.5526,
|
947 |
+
"step": 268
|
948 |
+
},
|
949 |
+
{
|
950 |
+
"epoch": 2.769230769230769,
|
951 |
+
"grad_norm": 0.6809393167495728,
|
952 |
+
"learning_rate": 7.226804123711341e-05,
|
953 |
+
"loss": 6.6546,
|
954 |
+
"step": 270
|
955 |
+
},
|
956 |
+
{
|
957 |
+
"epoch": 2.7897435897435896,
|
958 |
+
"grad_norm": 0.7717563509941101,
|
959 |
+
"learning_rate": 7.206185567010309e-05,
|
960 |
+
"loss": 6.3678,
|
961 |
+
"step": 272
|
962 |
+
},
|
963 |
+
{
|
964 |
+
"epoch": 2.81025641025641,
|
965 |
+
"grad_norm": 0.7897918820381165,
|
966 |
+
"learning_rate": 7.185567010309279e-05,
|
967 |
+
"loss": 6.2131,
|
968 |
+
"step": 274
|
969 |
+
},
|
970 |
+
{
|
971 |
+
"epoch": 2.830769230769231,
|
972 |
+
"grad_norm": 0.6473929286003113,
|
973 |
+
"learning_rate": 7.164948453608247e-05,
|
974 |
+
"loss": 6.6045,
|
975 |
+
"step": 276
|
976 |
+
},
|
977 |
+
{
|
978 |
+
"epoch": 2.8512820512820514,
|
979 |
+
"grad_norm": 0.7253041863441467,
|
980 |
+
"learning_rate": 7.144329896907217e-05,
|
981 |
+
"loss": 6.4355,
|
982 |
+
"step": 278
|
983 |
+
},
|
984 |
+
{
|
985 |
+
"epoch": 2.871794871794872,
|
986 |
+
"grad_norm": 0.6646418571472168,
|
987 |
+
"learning_rate": 7.123711340206187e-05,
|
988 |
+
"loss": 6.4636,
|
989 |
+
"step": 280
|
990 |
+
},
|
991 |
+
{
|
992 |
+
"epoch": 2.8923076923076922,
|
993 |
+
"grad_norm": 0.6046567559242249,
|
994 |
+
"learning_rate": 7.103092783505155e-05,
|
995 |
+
"loss": 6.3511,
|
996 |
+
"step": 282
|
997 |
+
},
|
998 |
+
{
|
999 |
+
"epoch": 2.9128205128205127,
|
1000 |
+
"grad_norm": 0.6396191120147705,
|
1001 |
+
"learning_rate": 7.082474226804125e-05,
|
1002 |
+
"loss": 6.5544,
|
1003 |
+
"step": 284
|
1004 |
+
},
|
1005 |
+
{
|
1006 |
+
"epoch": 2.9333333333333336,
|
1007 |
+
"grad_norm": 0.7889509797096252,
|
1008 |
+
"learning_rate": 7.061855670103093e-05,
|
1009 |
+
"loss": 6.5913,
|
1010 |
+
"step": 286
|
1011 |
+
},
|
1012 |
+
{
|
1013 |
+
"epoch": 2.953846153846154,
|
1014 |
+
"grad_norm": 0.7650195956230164,
|
1015 |
+
"learning_rate": 7.041237113402063e-05,
|
1016 |
+
"loss": 6.4351,
|
1017 |
+
"step": 288
|
1018 |
+
},
|
1019 |
+
{
|
1020 |
+
"epoch": 2.9743589743589745,
|
1021 |
+
"grad_norm": 0.7612701654434204,
|
1022 |
+
"learning_rate": 7.020618556701031e-05,
|
1023 |
+
"loss": 6.3414,
|
1024 |
+
"step": 290
|
1025 |
+
},
|
1026 |
+
{
|
1027 |
+
"epoch": 2.994871794871795,
|
1028 |
+
"grad_norm": 0.7275530695915222,
|
1029 |
+
"learning_rate": 7e-05,
|
1030 |
+
"loss": 6.5395,
|
1031 |
+
"step": 292
|
1032 |
+
},
|
1033 |
+
{
|
1034 |
+
"epoch": 3.0153846153846153,
|
1035 |
+
"grad_norm": 0.6664620637893677,
|
1036 |
+
"learning_rate": 6.979381443298969e-05,
|
1037 |
+
"loss": 6.1818,
|
1038 |
+
"step": 294
|
1039 |
+
},
|
1040 |
+
{
|
1041 |
+
"epoch": 3.0358974358974358,
|
1042 |
+
"grad_norm": 0.6935043334960938,
|
1043 |
+
"learning_rate": 6.958762886597939e-05,
|
1044 |
+
"loss": 6.2724,
|
1045 |
+
"step": 296
|
1046 |
+
},
|
1047 |
+
{
|
1048 |
+
"epoch": 3.056410256410256,
|
1049 |
+
"grad_norm": 0.691132664680481,
|
1050 |
+
"learning_rate": 6.938144329896907e-05,
|
1051 |
+
"loss": 6.3618,
|
1052 |
+
"step": 298
|
1053 |
+
},
|
1054 |
+
{
|
1055 |
+
"epoch": 3.076923076923077,
|
1056 |
+
"grad_norm": 0.6852958798408508,
|
1057 |
+
"learning_rate": 6.917525773195877e-05,
|
1058 |
+
"loss": 6.4248,
|
1059 |
+
"step": 300
|
1060 |
+
},
|
1061 |
+
{
|
1062 |
+
"epoch": 3.0974358974358975,
|
1063 |
+
"grad_norm": 0.7287474274635315,
|
1064 |
+
"learning_rate": 6.896907216494845e-05,
|
1065 |
+
"loss": 6.5134,
|
1066 |
+
"step": 302
|
1067 |
+
},
|
1068 |
+
{
|
1069 |
+
"epoch": 3.117948717948718,
|
1070 |
+
"grad_norm": 0.6913236379623413,
|
1071 |
+
"learning_rate": 6.876288659793815e-05,
|
1072 |
+
"loss": 6.5583,
|
1073 |
+
"step": 304
|
1074 |
+
},
|
1075 |
+
{
|
1076 |
+
"epoch": 3.1384615384615384,
|
1077 |
+
"grad_norm": 0.7557226419448853,
|
1078 |
+
"learning_rate": 6.855670103092783e-05,
|
1079 |
+
"loss": 6.3671,
|
1080 |
+
"step": 306
|
1081 |
+
},
|
1082 |
+
{
|
1083 |
+
"epoch": 3.158974358974359,
|
1084 |
+
"grad_norm": 0.673149049282074,
|
1085 |
+
"learning_rate": 6.835051546391753e-05,
|
1086 |
+
"loss": 6.463,
|
1087 |
+
"step": 308
|
1088 |
+
},
|
1089 |
+
{
|
1090 |
+
"epoch": 3.1794871794871793,
|
1091 |
+
"grad_norm": 0.7111195921897888,
|
1092 |
+
"learning_rate": 6.814432989690723e-05,
|
1093 |
+
"loss": 6.4296,
|
1094 |
+
"step": 310
|
1095 |
+
},
|
1096 |
+
{
|
1097 |
+
"epoch": 3.2,
|
1098 |
+
"grad_norm": 0.7462407946586609,
|
1099 |
+
"learning_rate": 6.793814432989691e-05,
|
1100 |
+
"loss": 6.3978,
|
1101 |
+
"step": 312
|
1102 |
+
},
|
1103 |
+
{
|
1104 |
+
"epoch": 3.2205128205128206,
|
1105 |
+
"grad_norm": 0.633978009223938,
|
1106 |
+
"learning_rate": 6.773195876288661e-05,
|
1107 |
+
"loss": 6.5069,
|
1108 |
+
"step": 314
|
1109 |
+
},
|
1110 |
+
{
|
1111 |
+
"epoch": 3.241025641025641,
|
1112 |
+
"grad_norm": 0.5375587940216064,
|
1113 |
+
"learning_rate": 6.752577319587629e-05,
|
1114 |
+
"loss": 6.5878,
|
1115 |
+
"step": 316
|
1116 |
+
},
|
1117 |
+
{
|
1118 |
+
"epoch": 3.2615384615384615,
|
1119 |
+
"grad_norm": 0.6872283816337585,
|
1120 |
+
"learning_rate": 6.731958762886599e-05,
|
1121 |
+
"loss": 6.5541,
|
1122 |
+
"step": 318
|
1123 |
+
},
|
1124 |
+
{
|
1125 |
+
"epoch": 3.282051282051282,
|
1126 |
+
"grad_norm": 0.7773441076278687,
|
1127 |
+
"learning_rate": 6.711340206185567e-05,
|
1128 |
+
"loss": 6.3327,
|
1129 |
+
"step": 320
|
1130 |
+
},
|
1131 |
+
{
|
1132 |
+
"epoch": 3.3025641025641024,
|
1133 |
+
"grad_norm": 0.9842550754547119,
|
1134 |
+
"learning_rate": 6.690721649484535e-05,
|
1135 |
+
"loss": 6.4966,
|
1136 |
+
"step": 322
|
1137 |
+
},
|
1138 |
+
{
|
1139 |
+
"epoch": 3.3230769230769233,
|
1140 |
+
"grad_norm": 0.5430623292922974,
|
1141 |
+
"learning_rate": 6.670103092783505e-05,
|
1142 |
+
"loss": 6.5381,
|
1143 |
+
"step": 324
|
1144 |
+
},
|
1145 |
+
{
|
1146 |
+
"epoch": 3.3435897435897437,
|
1147 |
+
"grad_norm": 0.6485368609428406,
|
1148 |
+
"learning_rate": 6.649484536082475e-05,
|
1149 |
+
"loss": 6.4691,
|
1150 |
+
"step": 326
|
1151 |
+
},
|
1152 |
+
{
|
1153 |
+
"epoch": 3.364102564102564,
|
1154 |
+
"grad_norm": 0.8230199217796326,
|
1155 |
+
"learning_rate": 6.628865979381443e-05,
|
1156 |
+
"loss": 6.3189,
|
1157 |
+
"step": 328
|
1158 |
+
},
|
1159 |
+
{
|
1160 |
+
"epoch": 3.3846153846153846,
|
1161 |
+
"grad_norm": 0.6916872262954712,
|
1162 |
+
"learning_rate": 6.608247422680413e-05,
|
1163 |
+
"loss": 6.3726,
|
1164 |
+
"step": 330
|
1165 |
+
},
|
1166 |
+
{
|
1167 |
+
"epoch": 3.405128205128205,
|
1168 |
+
"grad_norm": 0.6563546657562256,
|
1169 |
+
"learning_rate": 6.587628865979381e-05,
|
1170 |
+
"loss": 6.532,
|
1171 |
+
"step": 332
|
1172 |
+
},
|
1173 |
+
{
|
1174 |
+
"epoch": 3.4256410256410255,
|
1175 |
+
"grad_norm": 0.8116714954376221,
|
1176 |
+
"learning_rate": 6.567010309278351e-05,
|
1177 |
+
"loss": 6.3005,
|
1178 |
+
"step": 334
|
1179 |
+
},
|
1180 |
+
{
|
1181 |
+
"epoch": 3.4461538461538463,
|
1182 |
+
"grad_norm": 0.8624938726425171,
|
1183 |
+
"learning_rate": 6.546391752577319e-05,
|
1184 |
+
"loss": 6.2658,
|
1185 |
+
"step": 336
|
1186 |
+
},
|
1187 |
+
{
|
1188 |
+
"epoch": 3.466666666666667,
|
1189 |
+
"grad_norm": 0.8297551274299622,
|
1190 |
+
"learning_rate": 6.525773195876289e-05,
|
1191 |
+
"loss": 6.3995,
|
1192 |
+
"step": 338
|
1193 |
+
},
|
1194 |
+
{
|
1195 |
+
"epoch": 3.4871794871794872,
|
1196 |
+
"grad_norm": 0.6855469942092896,
|
1197 |
+
"learning_rate": 6.505154639175259e-05,
|
1198 |
+
"loss": 6.3061,
|
1199 |
+
"step": 340
|
1200 |
+
},
|
1201 |
+
{
|
1202 |
+
"epoch": 3.5076923076923077,
|
1203 |
+
"grad_norm": 0.7813113331794739,
|
1204 |
+
"learning_rate": 6.484536082474227e-05,
|
1205 |
+
"loss": 6.3236,
|
1206 |
+
"step": 342
|
1207 |
+
},
|
1208 |
+
{
|
1209 |
+
"epoch": 3.528205128205128,
|
1210 |
+
"grad_norm": 0.6506015062332153,
|
1211 |
+
"learning_rate": 6.463917525773197e-05,
|
1212 |
+
"loss": 6.4452,
|
1213 |
+
"step": 344
|
1214 |
+
},
|
1215 |
+
{
|
1216 |
+
"epoch": 3.5487179487179485,
|
1217 |
+
"grad_norm": 0.7242425084114075,
|
1218 |
+
"learning_rate": 6.443298969072165e-05,
|
1219 |
+
"loss": 6.4236,
|
1220 |
+
"step": 346
|
1221 |
+
},
|
1222 |
+
{
|
1223 |
+
"epoch": 3.569230769230769,
|
1224 |
+
"grad_norm": 0.7753481864929199,
|
1225 |
+
"learning_rate": 6.422680412371135e-05,
|
1226 |
+
"loss": 6.503,
|
1227 |
+
"step": 348
|
1228 |
+
},
|
1229 |
+
{
|
1230 |
+
"epoch": 3.58974358974359,
|
1231 |
+
"grad_norm": 0.8971739411354065,
|
1232 |
+
"learning_rate": 6.402061855670103e-05,
|
1233 |
+
"loss": 6.2226,
|
1234 |
+
"step": 350
|
1235 |
+
},
|
1236 |
+
{
|
1237 |
+
"epoch": 3.6102564102564103,
|
1238 |
+
"grad_norm": 0.719221293926239,
|
1239 |
+
"learning_rate": 6.381443298969071e-05,
|
1240 |
+
"loss": 6.4305,
|
1241 |
+
"step": 352
|
1242 |
+
},
|
1243 |
+
{
|
1244 |
+
"epoch": 3.6307692307692307,
|
1245 |
+
"grad_norm": 0.7092213034629822,
|
1246 |
+
"learning_rate": 6.360824742268041e-05,
|
1247 |
+
"loss": 6.5107,
|
1248 |
+
"step": 354
|
1249 |
+
},
|
1250 |
+
{
|
1251 |
+
"epoch": 3.651282051282051,
|
1252 |
+
"grad_norm": 0.69648277759552,
|
1253 |
+
"learning_rate": 6.340206185567011e-05,
|
1254 |
+
"loss": 6.1619,
|
1255 |
+
"step": 356
|
1256 |
+
},
|
1257 |
+
{
|
1258 |
+
"epoch": 3.6717948717948716,
|
1259 |
+
"grad_norm": 0.7983478903770447,
|
1260 |
+
"learning_rate": 6.319587628865979e-05,
|
1261 |
+
"loss": 6.2945,
|
1262 |
+
"step": 358
|
1263 |
+
},
|
1264 |
+
{
|
1265 |
+
"epoch": 3.6923076923076925,
|
1266 |
+
"grad_norm": 0.6972605586051941,
|
1267 |
+
"learning_rate": 6.298969072164949e-05,
|
1268 |
+
"loss": 6.3397,
|
1269 |
+
"step": 360
|
1270 |
+
},
|
1271 |
+
{
|
1272 |
+
"epoch": 3.712820512820513,
|
1273 |
+
"grad_norm": 0.7316955924034119,
|
1274 |
+
"learning_rate": 6.278350515463917e-05,
|
1275 |
+
"loss": 6.5642,
|
1276 |
+
"step": 362
|
1277 |
+
},
|
1278 |
+
{
|
1279 |
+
"epoch": 3.7333333333333334,
|
1280 |
+
"grad_norm": 0.819416344165802,
|
1281 |
+
"learning_rate": 6.257731958762887e-05,
|
1282 |
+
"loss": 6.5361,
|
1283 |
+
"step": 364
|
1284 |
+
},
|
1285 |
+
{
|
1286 |
+
"epoch": 3.753846153846154,
|
1287 |
+
"grad_norm": 0.7884580492973328,
|
1288 |
+
"learning_rate": 6.237113402061855e-05,
|
1289 |
+
"loss": 6.1174,
|
1290 |
+
"step": 366
|
1291 |
+
},
|
1292 |
+
{
|
1293 |
+
"epoch": 3.7743589743589743,
|
1294 |
+
"grad_norm": 0.7763500809669495,
|
1295 |
+
"learning_rate": 6.216494845360825e-05,
|
1296 |
+
"loss": 6.2631,
|
1297 |
+
"step": 368
|
1298 |
+
},
|
1299 |
+
{
|
1300 |
+
"epoch": 3.7948717948717947,
|
1301 |
+
"grad_norm": 0.8002706170082092,
|
1302 |
+
"learning_rate": 6.195876288659795e-05,
|
1303 |
+
"loss": 6.3999,
|
1304 |
+
"step": 370
|
1305 |
+
},
|
1306 |
+
{
|
1307 |
+
"epoch": 3.815384615384615,
|
1308 |
+
"grad_norm": 0.6830271482467651,
|
1309 |
+
"learning_rate": 6.175257731958763e-05,
|
1310 |
+
"loss": 6.4109,
|
1311 |
+
"step": 372
|
1312 |
+
},
|
1313 |
+
{
|
1314 |
+
"epoch": 3.835897435897436,
|
1315 |
+
"grad_norm": 0.8474668860435486,
|
1316 |
+
"learning_rate": 6.154639175257733e-05,
|
1317 |
+
"loss": 6.2914,
|
1318 |
+
"step": 374
|
1319 |
+
},
|
1320 |
+
{
|
1321 |
+
"epoch": 3.8564102564102565,
|
1322 |
+
"grad_norm": 0.7036705017089844,
|
1323 |
+
"learning_rate": 6.134020618556701e-05,
|
1324 |
+
"loss": 6.5863,
|
1325 |
+
"step": 376
|
1326 |
+
},
|
1327 |
+
{
|
1328 |
+
"epoch": 3.876923076923077,
|
1329 |
+
"grad_norm": 0.6333104968070984,
|
1330 |
+
"learning_rate": 6.113402061855671e-05,
|
1331 |
+
"loss": 6.5162,
|
1332 |
+
"step": 378
|
1333 |
+
},
|
1334 |
+
{
|
1335 |
+
"epoch": 3.8974358974358974,
|
1336 |
+
"grad_norm": 0.5705188512802124,
|
1337 |
+
"learning_rate": 6.092783505154639e-05,
|
1338 |
+
"loss": 6.3006,
|
1339 |
+
"step": 380
|
1340 |
+
},
|
1341 |
+
{
|
1342 |
+
"epoch": 3.917948717948718,
|
1343 |
+
"grad_norm": 0.634800136089325,
|
1344 |
+
"learning_rate": 6.072164948453608e-05,
|
1345 |
+
"loss": 6.4401,
|
1346 |
+
"step": 382
|
1347 |
+
},
|
1348 |
+
{
|
1349 |
+
"epoch": 3.9384615384615387,
|
1350 |
+
"grad_norm": 0.6740937232971191,
|
1351 |
+
"learning_rate": 6.051546391752577e-05,
|
1352 |
+
"loss": 6.698,
|
1353 |
+
"step": 384
|
1354 |
+
},
|
1355 |
+
{
|
1356 |
+
"epoch": 3.958974358974359,
|
1357 |
+
"grad_norm": 0.9393010139465332,
|
1358 |
+
"learning_rate": 6.030927835051546e-05,
|
1359 |
+
"loss": 6.1999,
|
1360 |
+
"step": 386
|
1361 |
+
},
|
1362 |
+
{
|
1363 |
+
"epoch": 3.9794871794871796,
|
1364 |
+
"grad_norm": 0.6255496740341187,
|
1365 |
+
"learning_rate": 6.010309278350515e-05,
|
1366 |
+
"loss": 6.3655,
|
1367 |
+
"step": 388
|
1368 |
+
},
|
1369 |
+
{
|
1370 |
+
"epoch": 4.0,
|
1371 |
+
"grad_norm": 0.6616120934486389,
|
1372 |
+
"learning_rate": 5.989690721649484e-05,
|
1373 |
+
"loss": 6.4234,
|
1374 |
+
"step": 390
|
1375 |
+
},
|
1376 |
+
{
|
1377 |
+
"epoch": 4.02051282051282,
|
1378 |
+
"grad_norm": 0.6440649032592773,
|
1379 |
+
"learning_rate": 5.969072164948454e-05,
|
1380 |
+
"loss": 6.3687,
|
1381 |
+
"step": 392
|
1382 |
+
},
|
1383 |
+
{
|
1384 |
+
"epoch": 4.041025641025641,
|
1385 |
+
"grad_norm": 0.6960980296134949,
|
1386 |
+
"learning_rate": 5.948453608247423e-05,
|
1387 |
+
"loss": 6.453,
|
1388 |
+
"step": 394
|
1389 |
+
},
|
1390 |
+
{
|
1391 |
+
"epoch": 4.061538461538461,
|
1392 |
+
"grad_norm": 0.8931025266647339,
|
1393 |
+
"learning_rate": 5.927835051546392e-05,
|
1394 |
+
"loss": 6.4502,
|
1395 |
+
"step": 396
|
1396 |
+
},
|
1397 |
+
{
|
1398 |
+
"epoch": 4.082051282051282,
|
1399 |
+
"grad_norm": 0.6508329510688782,
|
1400 |
+
"learning_rate": 5.907216494845361e-05,
|
1401 |
+
"loss": 6.4387,
|
1402 |
+
"step": 398
|
1403 |
+
},
|
1404 |
+
{
|
1405 |
+
"epoch": 4.102564102564102,
|
1406 |
+
"grad_norm": 0.7409791946411133,
|
1407 |
+
"learning_rate": 5.88659793814433e-05,
|
1408 |
+
"loss": 6.3901,
|
1409 |
+
"step": 400
|
1410 |
+
},
|
1411 |
+
{
|
1412 |
+
"epoch": 4.123076923076923,
|
1413 |
+
"grad_norm": 0.6797041893005371,
|
1414 |
+
"learning_rate": 5.8659793814433e-05,
|
1415 |
+
"loss": 6.3299,
|
1416 |
+
"step": 402
|
1417 |
+
},
|
1418 |
+
{
|
1419 |
+
"epoch": 4.143589743589744,
|
1420 |
+
"grad_norm": 0.7487124800682068,
|
1421 |
+
"learning_rate": 5.845360824742269e-05,
|
1422 |
+
"loss": 6.3846,
|
1423 |
+
"step": 404
|
1424 |
+
},
|
1425 |
+
{
|
1426 |
+
"epoch": 4.164102564102564,
|
1427 |
+
"grad_norm": 0.6995207071304321,
|
1428 |
+
"learning_rate": 5.824742268041238e-05,
|
1429 |
+
"loss": 6.4029,
|
1430 |
+
"step": 406
|
1431 |
+
},
|
1432 |
+
{
|
1433 |
+
"epoch": 4.184615384615385,
|
1434 |
+
"grad_norm": 0.920820951461792,
|
1435 |
+
"learning_rate": 5.804123711340207e-05,
|
1436 |
+
"loss": 6.398,
|
1437 |
+
"step": 408
|
1438 |
+
},
|
1439 |
+
{
|
1440 |
+
"epoch": 4.205128205128205,
|
1441 |
+
"grad_norm": 0.7857652306556702,
|
1442 |
+
"learning_rate": 5.783505154639176e-05,
|
1443 |
+
"loss": 6.2116,
|
1444 |
+
"step": 410
|
1445 |
+
},
|
1446 |
+
{
|
1447 |
+
"epoch": 4.225641025641026,
|
1448 |
+
"grad_norm": 0.655503511428833,
|
1449 |
+
"learning_rate": 5.762886597938144e-05,
|
1450 |
+
"loss": 6.4523,
|
1451 |
+
"step": 412
|
1452 |
+
},
|
1453 |
+
{
|
1454 |
+
"epoch": 4.246153846153846,
|
1455 |
+
"grad_norm": 0.7093043327331543,
|
1456 |
+
"learning_rate": 5.742268041237113e-05,
|
1457 |
+
"loss": 6.2784,
|
1458 |
+
"step": 414
|
1459 |
+
},
|
1460 |
+
{
|
1461 |
+
"epoch": 4.266666666666667,
|
1462 |
+
"grad_norm": 0.6597480177879333,
|
1463 |
+
"learning_rate": 5.721649484536082e-05,
|
1464 |
+
"loss": 6.4637,
|
1465 |
+
"step": 416
|
1466 |
+
},
|
1467 |
+
{
|
1468 |
+
"epoch": 4.287179487179487,
|
1469 |
+
"grad_norm": 0.8404126763343811,
|
1470 |
+
"learning_rate": 5.7010309278350514e-05,
|
1471 |
+
"loss": 6.226,
|
1472 |
+
"step": 418
|
1473 |
+
},
|
1474 |
+
{
|
1475 |
+
"epoch": 4.3076923076923075,
|
1476 |
+
"grad_norm": 0.7892007827758789,
|
1477 |
+
"learning_rate": 5.6804123711340204e-05,
|
1478 |
+
"loss": 6.3117,
|
1479 |
+
"step": 420
|
1480 |
+
},
|
1481 |
+
{
|
1482 |
+
"epoch": 4.328205128205128,
|
1483 |
+
"grad_norm": 0.7826732397079468,
|
1484 |
+
"learning_rate": 5.65979381443299e-05,
|
1485 |
+
"loss": 6.2826,
|
1486 |
+
"step": 422
|
1487 |
+
},
|
1488 |
+
{
|
1489 |
+
"epoch": 4.348717948717948,
|
1490 |
+
"grad_norm": 0.7233083248138428,
|
1491 |
+
"learning_rate": 5.639175257731959e-05,
|
1492 |
+
"loss": 6.3743,
|
1493 |
+
"step": 424
|
1494 |
+
},
|
1495 |
+
{
|
1496 |
+
"epoch": 4.36923076923077,
|
1497 |
+
"grad_norm": 0.6873509287834167,
|
1498 |
+
"learning_rate": 5.618556701030928e-05,
|
1499 |
+
"loss": 6.3368,
|
1500 |
+
"step": 426
|
1501 |
+
},
|
1502 |
+
{
|
1503 |
+
"epoch": 4.38974358974359,
|
1504 |
+
"grad_norm": 0.8165833353996277,
|
1505 |
+
"learning_rate": 5.597938144329897e-05,
|
1506 |
+
"loss": 6.4183,
|
1507 |
+
"step": 428
|
1508 |
+
},
|
1509 |
+
{
|
1510 |
+
"epoch": 4.410256410256411,
|
1511 |
+
"grad_norm": 0.7139970064163208,
|
1512 |
+
"learning_rate": 5.577319587628866e-05,
|
1513 |
+
"loss": 6.3427,
|
1514 |
+
"step": 430
|
1515 |
+
},
|
1516 |
+
{
|
1517 |
+
"epoch": 4.430769230769231,
|
1518 |
+
"grad_norm": 0.61057448387146,
|
1519 |
+
"learning_rate": 5.556701030927836e-05,
|
1520 |
+
"loss": 6.3747,
|
1521 |
+
"step": 432
|
1522 |
+
},
|
1523 |
+
{
|
1524 |
+
"epoch": 4.4512820512820515,
|
1525 |
+
"grad_norm": 0.8155067563056946,
|
1526 |
+
"learning_rate": 5.536082474226805e-05,
|
1527 |
+
"loss": 6.1935,
|
1528 |
+
"step": 434
|
1529 |
+
},
|
1530 |
+
{
|
1531 |
+
"epoch": 4.471794871794872,
|
1532 |
+
"grad_norm": 0.7220977544784546,
|
1533 |
+
"learning_rate": 5.515463917525774e-05,
|
1534 |
+
"loss": 6.4346,
|
1535 |
+
"step": 436
|
1536 |
+
},
|
1537 |
+
{
|
1538 |
+
"epoch": 4.492307692307692,
|
1539 |
+
"grad_norm": 0.6980003714561462,
|
1540 |
+
"learning_rate": 5.494845360824743e-05,
|
1541 |
+
"loss": 6.3355,
|
1542 |
+
"step": 438
|
1543 |
+
},
|
1544 |
+
{
|
1545 |
+
"epoch": 4.512820512820513,
|
1546 |
+
"grad_norm": 0.7350332736968994,
|
1547 |
+
"learning_rate": 5.474226804123712e-05,
|
1548 |
+
"loss": 6.2946,
|
1549 |
+
"step": 440
|
1550 |
+
},
|
1551 |
+
{
|
1552 |
+
"epoch": 4.533333333333333,
|
1553 |
+
"grad_norm": 0.6564266085624695,
|
1554 |
+
"learning_rate": 5.45360824742268e-05,
|
1555 |
+
"loss": 6.3044,
|
1556 |
+
"step": 442
|
1557 |
+
},
|
1558 |
+
{
|
1559 |
+
"epoch": 4.553846153846154,
|
1560 |
+
"grad_norm": 0.8935516476631165,
|
1561 |
+
"learning_rate": 5.4329896907216494e-05,
|
1562 |
+
"loss": 6.0784,
|
1563 |
+
"step": 444
|
1564 |
+
},
|
1565 |
+
{
|
1566 |
+
"epoch": 4.574358974358974,
|
1567 |
+
"grad_norm": 0.6622797846794128,
|
1568 |
+
"learning_rate": 5.4123711340206184e-05,
|
1569 |
+
"loss": 6.3122,
|
1570 |
+
"step": 446
|
1571 |
+
},
|
1572 |
+
{
|
1573 |
+
"epoch": 4.5948717948717945,
|
1574 |
+
"grad_norm": 0.6584466695785522,
|
1575 |
+
"learning_rate": 5.3917525773195874e-05,
|
1576 |
+
"loss": 6.4586,
|
1577 |
+
"step": 448
|
1578 |
+
},
|
1579 |
+
{
|
1580 |
+
"epoch": 4.615384615384615,
|
1581 |
+
"grad_norm": 0.7214877605438232,
|
1582 |
+
"learning_rate": 5.3711340206185564e-05,
|
1583 |
+
"loss": 6.4671,
|
1584 |
+
"step": 450
|
1585 |
+
},
|
1586 |
+
{
|
1587 |
+
"epoch": 4.635897435897435,
|
1588 |
+
"grad_norm": 0.6560644507408142,
|
1589 |
+
"learning_rate": 5.350515463917526e-05,
|
1590 |
+
"loss": 6.2472,
|
1591 |
+
"step": 452
|
1592 |
+
},
|
1593 |
+
{
|
1594 |
+
"epoch": 4.656410256410257,
|
1595 |
+
"grad_norm": 0.7048325538635254,
|
1596 |
+
"learning_rate": 5.329896907216495e-05,
|
1597 |
+
"loss": 6.5175,
|
1598 |
+
"step": 454
|
1599 |
+
},
|
1600 |
+
{
|
1601 |
+
"epoch": 4.676923076923077,
|
1602 |
+
"grad_norm": 0.6315538883209229,
|
1603 |
+
"learning_rate": 5.309278350515464e-05,
|
1604 |
+
"loss": 6.0901,
|
1605 |
+
"step": 456
|
1606 |
+
},
|
1607 |
+
{
|
1608 |
+
"epoch": 4.697435897435898,
|
1609 |
+
"grad_norm": 0.6339179277420044,
|
1610 |
+
"learning_rate": 5.288659793814433e-05,
|
1611 |
+
"loss": 6.3935,
|
1612 |
+
"step": 458
|
1613 |
+
},
|
1614 |
+
{
|
1615 |
+
"epoch": 4.717948717948718,
|
1616 |
+
"grad_norm": 0.8235088586807251,
|
1617 |
+
"learning_rate": 5.268041237113402e-05,
|
1618 |
+
"loss": 5.9898,
|
1619 |
+
"step": 460
|
1620 |
+
},
|
1621 |
+
{
|
1622 |
+
"epoch": 4.7384615384615385,
|
1623 |
+
"grad_norm": 0.7074024081230164,
|
1624 |
+
"learning_rate": 5.247422680412372e-05,
|
1625 |
+
"loss": 6.1804,
|
1626 |
+
"step": 462
|
1627 |
+
},
|
1628 |
+
{
|
1629 |
+
"epoch": 4.758974358974359,
|
1630 |
+
"grad_norm": 0.8095348477363586,
|
1631 |
+
"learning_rate": 5.226804123711341e-05,
|
1632 |
+
"loss": 6.2209,
|
1633 |
+
"step": 464
|
1634 |
+
},
|
1635 |
+
{
|
1636 |
+
"epoch": 4.779487179487179,
|
1637 |
+
"grad_norm": 0.641472578048706,
|
1638 |
+
"learning_rate": 5.20618556701031e-05,
|
1639 |
+
"loss": 6.3102,
|
1640 |
+
"step": 466
|
1641 |
+
},
|
1642 |
+
{
|
1643 |
+
"epoch": 4.8,
|
1644 |
+
"grad_norm": 0.7063437700271606,
|
1645 |
+
"learning_rate": 5.185567010309279e-05,
|
1646 |
+
"loss": 6.2348,
|
1647 |
+
"step": 468
|
1648 |
+
},
|
1649 |
+
{
|
1650 |
+
"epoch": 4.82051282051282,
|
1651 |
+
"grad_norm": 0.7544019818305969,
|
1652 |
+
"learning_rate": 5.164948453608248e-05,
|
1653 |
+
"loss": 6.3763,
|
1654 |
+
"step": 470
|
1655 |
+
},
|
1656 |
+
{
|
1657 |
+
"epoch": 4.841025641025641,
|
1658 |
+
"grad_norm": 0.827674925327301,
|
1659 |
+
"learning_rate": 5.1443298969072164e-05,
|
1660 |
+
"loss": 6.0069,
|
1661 |
+
"step": 472
|
1662 |
+
},
|
1663 |
+
{
|
1664 |
+
"epoch": 4.861538461538462,
|
1665 |
+
"grad_norm": 0.8011050224304199,
|
1666 |
+
"learning_rate": 5.1237113402061854e-05,
|
1667 |
+
"loss": 6.257,
|
1668 |
+
"step": 474
|
1669 |
+
},
|
1670 |
+
{
|
1671 |
+
"epoch": 4.8820512820512825,
|
1672 |
+
"grad_norm": 0.6386663913726807,
|
1673 |
+
"learning_rate": 5.1030927835051544e-05,
|
1674 |
+
"loss": 6.3936,
|
1675 |
+
"step": 476
|
1676 |
+
},
|
1677 |
+
{
|
1678 |
+
"epoch": 4.902564102564103,
|
1679 |
+
"grad_norm": 0.7228955030441284,
|
1680 |
+
"learning_rate": 5.0824742268041235e-05,
|
1681 |
+
"loss": 6.4586,
|
1682 |
+
"step": 478
|
1683 |
+
},
|
1684 |
+
{
|
1685 |
+
"epoch": 4.923076923076923,
|
1686 |
+
"grad_norm": 0.6252913475036621,
|
1687 |
+
"learning_rate": 5.0618556701030925e-05,
|
1688 |
+
"loss": 6.3536,
|
1689 |
+
"step": 480
|
1690 |
+
},
|
1691 |
+
{
|
1692 |
+
"epoch": 4.943589743589744,
|
1693 |
+
"grad_norm": 0.7947326898574829,
|
1694 |
+
"learning_rate": 5.041237113402062e-05,
|
1695 |
+
"loss": 6.3856,
|
1696 |
+
"step": 482
|
1697 |
+
},
|
1698 |
+
{
|
1699 |
+
"epoch": 4.964102564102564,
|
1700 |
+
"grad_norm": 0.6296289563179016,
|
1701 |
+
"learning_rate": 5.020618556701031e-05,
|
1702 |
+
"loss": 6.4445,
|
1703 |
+
"step": 484
|
1704 |
+
},
|
1705 |
+
{
|
1706 |
+
"epoch": 4.984615384615385,
|
1707 |
+
"grad_norm": 0.6455388069152832,
|
1708 |
+
"learning_rate": 5e-05,
|
1709 |
+
"loss": 6.4503,
|
1710 |
+
"step": 486
|
1711 |
+
},
|
1712 |
+
{
|
1713 |
+
"epoch": 5.005128205128205,
|
1714 |
+
"grad_norm": 0.7692280411720276,
|
1715 |
+
"learning_rate": 4.979381443298969e-05,
|
1716 |
+
"loss": 6.4011,
|
1717 |
+
"step": 488
|
1718 |
+
},
|
1719 |
+
{
|
1720 |
+
"epoch": 5.0256410256410255,
|
1721 |
+
"grad_norm": 0.6310716867446899,
|
1722 |
+
"learning_rate": 4.958762886597938e-05,
|
1723 |
+
"loss": 6.3749,
|
1724 |
+
"step": 490
|
1725 |
+
},
|
1726 |
+
{
|
1727 |
+
"epoch": 5.046153846153846,
|
1728 |
+
"grad_norm": 0.8823466897010803,
|
1729 |
+
"learning_rate": 4.938144329896908e-05,
|
1730 |
+
"loss": 6.2891,
|
1731 |
+
"step": 492
|
1732 |
+
},
|
1733 |
+
{
|
1734 |
+
"epoch": 5.066666666666666,
|
1735 |
+
"grad_norm": 0.7742350101470947,
|
1736 |
+
"learning_rate": 4.9175257731958763e-05,
|
1737 |
+
"loss": 6.1306,
|
1738 |
+
"step": 494
|
1739 |
+
},
|
1740 |
+
{
|
1741 |
+
"epoch": 5.087179487179487,
|
1742 |
+
"grad_norm": 0.6190592050552368,
|
1743 |
+
"learning_rate": 4.8969072164948454e-05,
|
1744 |
+
"loss": 6.3312,
|
1745 |
+
"step": 496
|
1746 |
+
},
|
1747 |
+
{
|
1748 |
+
"epoch": 5.107692307692307,
|
1749 |
+
"grad_norm": 0.7244901657104492,
|
1750 |
+
"learning_rate": 4.8762886597938144e-05,
|
1751 |
+
"loss": 6.3803,
|
1752 |
+
"step": 498
|
1753 |
+
},
|
1754 |
+
{
|
1755 |
+
"epoch": 5.128205128205128,
|
1756 |
+
"grad_norm": 0.7023468017578125,
|
1757 |
+
"learning_rate": 4.8556701030927834e-05,
|
1758 |
+
"loss": 6.2492,
|
1759 |
+
"step": 500
|
1760 |
+
},
|
1761 |
+
{
|
1762 |
+
"epoch": 5.148717948717949,
|
1763 |
+
"grad_norm": 0.7719900608062744,
|
1764 |
+
"learning_rate": 4.835051546391753e-05,
|
1765 |
+
"loss": 6.1033,
|
1766 |
+
"step": 502
|
1767 |
+
},
|
1768 |
+
{
|
1769 |
+
"epoch": 5.1692307692307695,
|
1770 |
+
"grad_norm": 0.8199257254600525,
|
1771 |
+
"learning_rate": 4.814432989690722e-05,
|
1772 |
+
"loss": 6.1981,
|
1773 |
+
"step": 504
|
1774 |
+
},
|
1775 |
+
{
|
1776 |
+
"epoch": 5.18974358974359,
|
1777 |
+
"grad_norm": 0.8238953351974487,
|
1778 |
+
"learning_rate": 4.793814432989691e-05,
|
1779 |
+
"loss": 6.2671,
|
1780 |
+
"step": 506
|
1781 |
+
},
|
1782 |
+
{
|
1783 |
+
"epoch": 5.21025641025641,
|
1784 |
+
"grad_norm": 0.5926157832145691,
|
1785 |
+
"learning_rate": 4.77319587628866e-05,
|
1786 |
+
"loss": 6.3837,
|
1787 |
+
"step": 508
|
1788 |
+
},
|
1789 |
+
{
|
1790 |
+
"epoch": 5.230769230769231,
|
1791 |
+
"grad_norm": 0.6979117393493652,
|
1792 |
+
"learning_rate": 4.7525773195876285e-05,
|
1793 |
+
"loss": 6.2836,
|
1794 |
+
"step": 510
|
1795 |
+
},
|
1796 |
+
{
|
1797 |
+
"epoch": 5.251282051282051,
|
1798 |
+
"grad_norm": 0.6661803126335144,
|
1799 |
+
"learning_rate": 4.731958762886598e-05,
|
1800 |
+
"loss": 6.1655,
|
1801 |
+
"step": 512
|
1802 |
+
},
|
1803 |
+
{
|
1804 |
+
"epoch": 5.271794871794872,
|
1805 |
+
"grad_norm": 0.6727478504180908,
|
1806 |
+
"learning_rate": 4.711340206185567e-05,
|
1807 |
+
"loss": 6.0956,
|
1808 |
+
"step": 514
|
1809 |
+
},
|
1810 |
+
{
|
1811 |
+
"epoch": 5.292307692307692,
|
1812 |
+
"grad_norm": 0.665337085723877,
|
1813 |
+
"learning_rate": 4.690721649484536e-05,
|
1814 |
+
"loss": 6.3724,
|
1815 |
+
"step": 516
|
1816 |
+
},
|
1817 |
+
{
|
1818 |
+
"epoch": 5.312820512820513,
|
1819 |
+
"grad_norm": 0.702816903591156,
|
1820 |
+
"learning_rate": 4.670103092783505e-05,
|
1821 |
+
"loss": 6.2241,
|
1822 |
+
"step": 518
|
1823 |
+
},
|
1824 |
+
{
|
1825 |
+
"epoch": 5.333333333333333,
|
1826 |
+
"grad_norm": 0.6713479161262512,
|
1827 |
+
"learning_rate": 4.6494845360824743e-05,
|
1828 |
+
"loss": 6.3077,
|
1829 |
+
"step": 520
|
1830 |
+
},
|
1831 |
+
{
|
1832 |
+
"epoch": 5.3538461538461535,
|
1833 |
+
"grad_norm": 0.6276135444641113,
|
1834 |
+
"learning_rate": 4.628865979381444e-05,
|
1835 |
+
"loss": 6.4411,
|
1836 |
+
"step": 522
|
1837 |
+
},
|
1838 |
+
{
|
1839 |
+
"epoch": 5.374358974358975,
|
1840 |
+
"grad_norm": 0.7353730201721191,
|
1841 |
+
"learning_rate": 4.6082474226804124e-05,
|
1842 |
+
"loss": 6.1735,
|
1843 |
+
"step": 524
|
1844 |
+
},
|
1845 |
+
{
|
1846 |
+
"epoch": 5.394871794871795,
|
1847 |
+
"grad_norm": 0.7393136620521545,
|
1848 |
+
"learning_rate": 4.5876288659793814e-05,
|
1849 |
+
"loss": 6.413,
|
1850 |
+
"step": 526
|
1851 |
+
},
|
1852 |
+
{
|
1853 |
+
"epoch": 5.415384615384616,
|
1854 |
+
"grad_norm": 0.5835318565368652,
|
1855 |
+
"learning_rate": 4.5670103092783504e-05,
|
1856 |
+
"loss": 6.4997,
|
1857 |
+
"step": 528
|
1858 |
+
},
|
1859 |
+
{
|
1860 |
+
"epoch": 5.435897435897436,
|
1861 |
+
"grad_norm": 0.6543262600898743,
|
1862 |
+
"learning_rate": 4.5463917525773195e-05,
|
1863 |
+
"loss": 6.3292,
|
1864 |
+
"step": 530
|
1865 |
+
},
|
1866 |
+
{
|
1867 |
+
"epoch": 5.456410256410257,
|
1868 |
+
"grad_norm": 0.7140960693359375,
|
1869 |
+
"learning_rate": 4.525773195876289e-05,
|
1870 |
+
"loss": 6.3267,
|
1871 |
+
"step": 532
|
1872 |
+
},
|
1873 |
+
{
|
1874 |
+
"epoch": 5.476923076923077,
|
1875 |
+
"grad_norm": 0.6978656649589539,
|
1876 |
+
"learning_rate": 4.505154639175258e-05,
|
1877 |
+
"loss": 6.4343,
|
1878 |
+
"step": 534
|
1879 |
+
},
|
1880 |
+
{
|
1881 |
+
"epoch": 5.4974358974358974,
|
1882 |
+
"grad_norm": 0.7488942742347717,
|
1883 |
+
"learning_rate": 4.484536082474227e-05,
|
1884 |
+
"loss": 6.3557,
|
1885 |
+
"step": 536
|
1886 |
+
},
|
1887 |
+
{
|
1888 |
+
"epoch": 5.517948717948718,
|
1889 |
+
"grad_norm": 0.753161609172821,
|
1890 |
+
"learning_rate": 4.463917525773196e-05,
|
1891 |
+
"loss": 6.2488,
|
1892 |
+
"step": 538
|
1893 |
+
},
|
1894 |
+
{
|
1895 |
+
"epoch": 5.538461538461538,
|
1896 |
+
"grad_norm": 0.7196393609046936,
|
1897 |
+
"learning_rate": 4.4432989690721646e-05,
|
1898 |
+
"loss": 6.2132,
|
1899 |
+
"step": 540
|
1900 |
+
},
|
1901 |
+
{
|
1902 |
+
"epoch": 5.558974358974359,
|
1903 |
+
"grad_norm": 0.681937038898468,
|
1904 |
+
"learning_rate": 4.422680412371134e-05,
|
1905 |
+
"loss": 6.1831,
|
1906 |
+
"step": 542
|
1907 |
+
},
|
1908 |
+
{
|
1909 |
+
"epoch": 5.579487179487179,
|
1910 |
+
"grad_norm": 0.6542847752571106,
|
1911 |
+
"learning_rate": 4.402061855670103e-05,
|
1912 |
+
"loss": 6.3156,
|
1913 |
+
"step": 544
|
1914 |
+
},
|
1915 |
+
{
|
1916 |
+
"epoch": 5.6,
|
1917 |
+
"grad_norm": 0.7364455461502075,
|
1918 |
+
"learning_rate": 4.3814432989690723e-05,
|
1919 |
+
"loss": 6.3007,
|
1920 |
+
"step": 546
|
1921 |
+
},
|
1922 |
+
{
|
1923 |
+
"epoch": 5.62051282051282,
|
1924 |
+
"grad_norm": 0.7742012739181519,
|
1925 |
+
"learning_rate": 4.3608247422680414e-05,
|
1926 |
+
"loss": 5.9234,
|
1927 |
+
"step": 548
|
1928 |
+
},
|
1929 |
+
{
|
1930 |
+
"epoch": 5.641025641025641,
|
1931 |
+
"grad_norm": 0.7302541136741638,
|
1932 |
+
"learning_rate": 4.3402061855670104e-05,
|
1933 |
+
"loss": 6.3115,
|
1934 |
+
"step": 550
|
1935 |
+
},
|
1936 |
+
{
|
1937 |
+
"epoch": 5.661538461538462,
|
1938 |
+
"grad_norm": 0.7138387560844421,
|
1939 |
+
"learning_rate": 4.31958762886598e-05,
|
1940 |
+
"loss": 6.1719,
|
1941 |
+
"step": 552
|
1942 |
+
},
|
1943 |
+
{
|
1944 |
+
"epoch": 5.682051282051282,
|
1945 |
+
"grad_norm": 0.6778889298439026,
|
1946 |
+
"learning_rate": 4.298969072164949e-05,
|
1947 |
+
"loss": 6.1886,
|
1948 |
+
"step": 554
|
1949 |
+
},
|
1950 |
+
{
|
1951 |
+
"epoch": 5.702564102564103,
|
1952 |
+
"grad_norm": 0.7845616936683655,
|
1953 |
+
"learning_rate": 4.2783505154639175e-05,
|
1954 |
+
"loss": 6.1925,
|
1955 |
+
"step": 556
|
1956 |
+
},
|
1957 |
+
{
|
1958 |
+
"epoch": 5.723076923076923,
|
1959 |
+
"grad_norm": 0.780630886554718,
|
1960 |
+
"learning_rate": 4.2577319587628865e-05,
|
1961 |
+
"loss": 6.3325,
|
1962 |
+
"step": 558
|
1963 |
+
},
|
1964 |
+
{
|
1965 |
+
"epoch": 5.743589743589744,
|
1966 |
+
"grad_norm": 0.630280613899231,
|
1967 |
+
"learning_rate": 4.2371134020618555e-05,
|
1968 |
+
"loss": 6.3334,
|
1969 |
+
"step": 560
|
1970 |
+
},
|
1971 |
+
{
|
1972 |
+
"epoch": 5.764102564102564,
|
1973 |
+
"grad_norm": 0.7802500128746033,
|
1974 |
+
"learning_rate": 4.216494845360825e-05,
|
1975 |
+
"loss": 6.2872,
|
1976 |
+
"step": 562
|
1977 |
+
},
|
1978 |
+
{
|
1979 |
+
"epoch": 5.7846153846153845,
|
1980 |
+
"grad_norm": 0.6819648146629333,
|
1981 |
+
"learning_rate": 4.195876288659794e-05,
|
1982 |
+
"loss": 6.3133,
|
1983 |
+
"step": 564
|
1984 |
+
},
|
1985 |
+
{
|
1986 |
+
"epoch": 5.805128205128205,
|
1987 |
+
"grad_norm": 0.7848822474479675,
|
1988 |
+
"learning_rate": 4.175257731958763e-05,
|
1989 |
+
"loss": 6.4798,
|
1990 |
+
"step": 566
|
1991 |
+
},
|
1992 |
+
{
|
1993 |
+
"epoch": 5.825641025641025,
|
1994 |
+
"grad_norm": 0.7532023787498474,
|
1995 |
+
"learning_rate": 4.154639175257732e-05,
|
1996 |
+
"loss": 6.1145,
|
1997 |
+
"step": 568
|
1998 |
+
},
|
1999 |
+
{
|
2000 |
+
"epoch": 5.846153846153846,
|
2001 |
+
"grad_norm": 0.679144561290741,
|
2002 |
+
"learning_rate": 4.1340206185567006e-05,
|
2003 |
+
"loss": 6.2509,
|
2004 |
+
"step": 570
|
2005 |
+
},
|
2006 |
+
{
|
2007 |
+
"epoch": 5.866666666666667,
|
2008 |
+
"grad_norm": 0.7792911529541016,
|
2009 |
+
"learning_rate": 4.1134020618556704e-05,
|
2010 |
+
"loss": 6.1544,
|
2011 |
+
"step": 572
|
2012 |
+
},
|
2013 |
+
{
|
2014 |
+
"epoch": 5.887179487179488,
|
2015 |
+
"grad_norm": 0.7862560153007507,
|
2016 |
+
"learning_rate": 4.0927835051546394e-05,
|
2017 |
+
"loss": 6.2661,
|
2018 |
+
"step": 574
|
2019 |
+
},
|
2020 |
+
{
|
2021 |
+
"epoch": 5.907692307692308,
|
2022 |
+
"grad_norm": 0.6902645826339722,
|
2023 |
+
"learning_rate": 4.0721649484536084e-05,
|
2024 |
+
"loss": 6.3423,
|
2025 |
+
"step": 576
|
2026 |
+
},
|
2027 |
+
{
|
2028 |
+
"epoch": 5.9282051282051285,
|
2029 |
+
"grad_norm": 0.6430610418319702,
|
2030 |
+
"learning_rate": 4.0515463917525774e-05,
|
2031 |
+
"loss": 6.3058,
|
2032 |
+
"step": 578
|
2033 |
+
},
|
2034 |
+
{
|
2035 |
+
"epoch": 5.948717948717949,
|
2036 |
+
"grad_norm": 0.7760281562805176,
|
2037 |
+
"learning_rate": 4.0309278350515465e-05,
|
2038 |
+
"loss": 6.3007,
|
2039 |
+
"step": 580
|
2040 |
+
},
|
2041 |
+
{
|
2042 |
+
"epoch": 5.969230769230769,
|
2043 |
+
"grad_norm": 0.7323822975158691,
|
2044 |
+
"learning_rate": 4.010309278350516e-05,
|
2045 |
+
"loss": 6.3434,
|
2046 |
+
"step": 582
|
2047 |
+
},
|
2048 |
+
{
|
2049 |
+
"epoch": 5.98974358974359,
|
2050 |
+
"grad_norm": 0.7060137391090393,
|
2051 |
+
"learning_rate": 3.989690721649485e-05,
|
2052 |
+
"loss": 6.2571,
|
2053 |
+
"step": 584
|
2054 |
+
},
|
2055 |
+
{
|
2056 |
+
"epoch": 6.01025641025641,
|
2057 |
+
"grad_norm": 0.7796257734298706,
|
2058 |
+
"learning_rate": 3.9690721649484535e-05,
|
2059 |
+
"loss": 6.1228,
|
2060 |
+
"step": 586
|
2061 |
+
},
|
2062 |
+
{
|
2063 |
+
"epoch": 6.030769230769231,
|
2064 |
+
"grad_norm": 0.8794940710067749,
|
2065 |
+
"learning_rate": 3.9484536082474226e-05,
|
2066 |
+
"loss": 6.3449,
|
2067 |
+
"step": 588
|
2068 |
+
},
|
2069 |
+
{
|
2070 |
+
"epoch": 6.051282051282051,
|
2071 |
+
"grad_norm": 0.6452521681785583,
|
2072 |
+
"learning_rate": 3.9278350515463916e-05,
|
2073 |
+
"loss": 6.0527,
|
2074 |
+
"step": 590
|
2075 |
+
},
|
2076 |
+
{
|
2077 |
+
"epoch": 6.0717948717948715,
|
2078 |
+
"grad_norm": 0.7071980834007263,
|
2079 |
+
"learning_rate": 3.907216494845361e-05,
|
2080 |
+
"loss": 6.1828,
|
2081 |
+
"step": 592
|
2082 |
+
},
|
2083 |
+
{
|
2084 |
+
"epoch": 6.092307692307692,
|
2085 |
+
"grad_norm": 0.6614767909049988,
|
2086 |
+
"learning_rate": 3.88659793814433e-05,
|
2087 |
+
"loss": 6.2702,
|
2088 |
+
"step": 594
|
2089 |
+
},
|
2090 |
+
{
|
2091 |
+
"epoch": 6.112820512820512,
|
2092 |
+
"grad_norm": 0.6533995866775513,
|
2093 |
+
"learning_rate": 3.865979381443299e-05,
|
2094 |
+
"loss": 6.2421,
|
2095 |
+
"step": 596
|
2096 |
+
},
|
2097 |
+
{
|
2098 |
+
"epoch": 6.133333333333334,
|
2099 |
+
"grad_norm": 0.7907997965812683,
|
2100 |
+
"learning_rate": 3.8453608247422684e-05,
|
2101 |
+
"loss": 6.4282,
|
2102 |
+
"step": 598
|
2103 |
+
},
|
2104 |
+
{
|
2105 |
+
"epoch": 6.153846153846154,
|
2106 |
+
"grad_norm": 0.709160327911377,
|
2107 |
+
"learning_rate": 3.824742268041237e-05,
|
2108 |
+
"loss": 6.0814,
|
2109 |
+
"step": 600
|
2110 |
+
},
|
2111 |
+
{
|
2112 |
+
"epoch": 6.174358974358975,
|
2113 |
+
"grad_norm": 0.8180028796195984,
|
2114 |
+
"learning_rate": 3.8041237113402064e-05,
|
2115 |
+
"loss": 6.3014,
|
2116 |
+
"step": 602
|
2117 |
+
},
|
2118 |
+
{
|
2119 |
+
"epoch": 6.194871794871795,
|
2120 |
+
"grad_norm": 0.6765635013580322,
|
2121 |
+
"learning_rate": 3.7835051546391754e-05,
|
2122 |
+
"loss": 6.3844,
|
2123 |
+
"step": 604
|
2124 |
+
},
|
2125 |
+
{
|
2126 |
+
"epoch": 6.2153846153846155,
|
2127 |
+
"grad_norm": 0.6009554862976074,
|
2128 |
+
"learning_rate": 3.7628865979381445e-05,
|
2129 |
+
"loss": 6.1951,
|
2130 |
+
"step": 606
|
2131 |
+
},
|
2132 |
+
{
|
2133 |
+
"epoch": 6.235897435897436,
|
2134 |
+
"grad_norm": 0.708351731300354,
|
2135 |
+
"learning_rate": 3.7422680412371135e-05,
|
2136 |
+
"loss": 6.4139,
|
2137 |
+
"step": 608
|
2138 |
+
},
|
2139 |
+
{
|
2140 |
+
"epoch": 6.256410256410256,
|
2141 |
+
"grad_norm": 0.7396306395530701,
|
2142 |
+
"learning_rate": 3.7216494845360825e-05,
|
2143 |
+
"loss": 6.194,
|
2144 |
+
"step": 610
|
2145 |
+
},
|
2146 |
+
{
|
2147 |
+
"epoch": 6.276923076923077,
|
2148 |
+
"grad_norm": 0.772692859172821,
|
2149 |
+
"learning_rate": 3.701030927835052e-05,
|
2150 |
+
"loss": 6.2968,
|
2151 |
+
"step": 612
|
2152 |
+
},
|
2153 |
+
{
|
2154 |
+
"epoch": 6.297435897435897,
|
2155 |
+
"grad_norm": 0.8002333641052246,
|
2156 |
+
"learning_rate": 3.680412371134021e-05,
|
2157 |
+
"loss": 6.2677,
|
2158 |
+
"step": 614
|
2159 |
+
},
|
2160 |
+
{
|
2161 |
+
"epoch": 6.317948717948718,
|
2162 |
+
"grad_norm": 1.078007698059082,
|
2163 |
+
"learning_rate": 3.6597938144329896e-05,
|
2164 |
+
"loss": 6.2674,
|
2165 |
+
"step": 616
|
2166 |
+
},
|
2167 |
+
{
|
2168 |
+
"epoch": 6.338461538461538,
|
2169 |
+
"grad_norm": 0.6645541787147522,
|
2170 |
+
"learning_rate": 3.6391752577319586e-05,
|
2171 |
+
"loss": 6.1999,
|
2172 |
+
"step": 618
|
2173 |
+
},
|
2174 |
+
{
|
2175 |
+
"epoch": 6.358974358974359,
|
2176 |
+
"grad_norm": 0.6708665490150452,
|
2177 |
+
"learning_rate": 3.6185567010309276e-05,
|
2178 |
+
"loss": 6.4461,
|
2179 |
+
"step": 620
|
2180 |
+
},
|
2181 |
+
{
|
2182 |
+
"epoch": 6.37948717948718,
|
2183 |
+
"grad_norm": 0.6594524383544922,
|
2184 |
+
"learning_rate": 3.597938144329897e-05,
|
2185 |
+
"loss": 6.2954,
|
2186 |
+
"step": 622
|
2187 |
+
},
|
2188 |
+
{
|
2189 |
+
"epoch": 6.4,
|
2190 |
+
"grad_norm": 0.9436811804771423,
|
2191 |
+
"learning_rate": 3.5773195876288664e-05,
|
2192 |
+
"loss": 6.1473,
|
2193 |
+
"step": 624
|
2194 |
+
},
|
2195 |
+
{
|
2196 |
+
"epoch": 6.420512820512821,
|
2197 |
+
"grad_norm": 0.7767890691757202,
|
2198 |
+
"learning_rate": 3.5567010309278354e-05,
|
2199 |
+
"loss": 6.2695,
|
2200 |
+
"step": 626
|
2201 |
+
},
|
2202 |
+
{
|
2203 |
+
"epoch": 6.441025641025641,
|
2204 |
+
"grad_norm": 0.8223767876625061,
|
2205 |
+
"learning_rate": 3.5360824742268044e-05,
|
2206 |
+
"loss": 6.2998,
|
2207 |
+
"step": 628
|
2208 |
+
},
|
2209 |
+
{
|
2210 |
+
"epoch": 6.461538461538462,
|
2211 |
+
"grad_norm": 0.753822922706604,
|
2212 |
+
"learning_rate": 3.515463917525773e-05,
|
2213 |
+
"loss": 6.1723,
|
2214 |
+
"step": 630
|
2215 |
+
},
|
2216 |
+
{
|
2217 |
+
"epoch": 6.482051282051282,
|
2218 |
+
"grad_norm": 0.7371943593025208,
|
2219 |
+
"learning_rate": 3.4948453608247425e-05,
|
2220 |
+
"loss": 6.2228,
|
2221 |
+
"step": 632
|
2222 |
+
},
|
2223 |
+
{
|
2224 |
+
"epoch": 6.5025641025641026,
|
2225 |
+
"grad_norm": 0.6721214056015015,
|
2226 |
+
"learning_rate": 3.4742268041237115e-05,
|
2227 |
+
"loss": 6.06,
|
2228 |
+
"step": 634
|
2229 |
+
},
|
2230 |
+
{
|
2231 |
+
"epoch": 6.523076923076923,
|
2232 |
+
"grad_norm": 0.7266871333122253,
|
2233 |
+
"learning_rate": 3.4536082474226805e-05,
|
2234 |
+
"loss": 6.3087,
|
2235 |
+
"step": 636
|
2236 |
+
},
|
2237 |
+
{
|
2238 |
+
"epoch": 6.543589743589743,
|
2239 |
+
"grad_norm": 0.7427794933319092,
|
2240 |
+
"learning_rate": 3.4329896907216495e-05,
|
2241 |
+
"loss": 6.2757,
|
2242 |
+
"step": 638
|
2243 |
+
},
|
2244 |
+
{
|
2245 |
+
"epoch": 6.564102564102564,
|
2246 |
+
"grad_norm": 0.7372191548347473,
|
2247 |
+
"learning_rate": 3.4123711340206186e-05,
|
2248 |
+
"loss": 6.1809,
|
2249 |
+
"step": 640
|
2250 |
+
},
|
2251 |
+
{
|
2252 |
+
"epoch": 6.584615384615384,
|
2253 |
+
"grad_norm": 0.773668646812439,
|
2254 |
+
"learning_rate": 3.391752577319588e-05,
|
2255 |
+
"loss": 6.0375,
|
2256 |
+
"step": 642
|
2257 |
+
},
|
2258 |
+
{
|
2259 |
+
"epoch": 6.605128205128205,
|
2260 |
+
"grad_norm": 0.8196036219596863,
|
2261 |
+
"learning_rate": 3.371134020618557e-05,
|
2262 |
+
"loss": 6.3719,
|
2263 |
+
"step": 644
|
2264 |
+
},
|
2265 |
+
{
|
2266 |
+
"epoch": 6.625641025641025,
|
2267 |
+
"grad_norm": 0.6300514936447144,
|
2268 |
+
"learning_rate": 3.3505154639175256e-05,
|
2269 |
+
"loss": 6.338,
|
2270 |
+
"step": 646
|
2271 |
+
},
|
2272 |
+
{
|
2273 |
+
"epoch": 6.6461538461538465,
|
2274 |
+
"grad_norm": 0.5757874846458435,
|
2275 |
+
"learning_rate": 3.3298969072164947e-05,
|
2276 |
+
"loss": 6.4443,
|
2277 |
+
"step": 648
|
2278 |
+
},
|
2279 |
+
{
|
2280 |
+
"epoch": 6.666666666666667,
|
2281 |
+
"grad_norm": 0.8658593893051147,
|
2282 |
+
"learning_rate": 3.309278350515464e-05,
|
2283 |
+
"loss": 6.1216,
|
2284 |
+
"step": 650
|
2285 |
+
},
|
2286 |
+
{
|
2287 |
+
"epoch": 6.687179487179487,
|
2288 |
+
"grad_norm": 0.7729750275611877,
|
2289 |
+
"learning_rate": 3.2886597938144334e-05,
|
2290 |
+
"loss": 5.9493,
|
2291 |
+
"step": 652
|
2292 |
+
},
|
2293 |
+
{
|
2294 |
+
"epoch": 6.707692307692308,
|
2295 |
+
"grad_norm": 0.7950991988182068,
|
2296 |
+
"learning_rate": 3.2680412371134024e-05,
|
2297 |
+
"loss": 6.1614,
|
2298 |
+
"step": 654
|
2299 |
+
},
|
2300 |
+
{
|
2301 |
+
"epoch": 6.728205128205128,
|
2302 |
+
"grad_norm": 0.7309786081314087,
|
2303 |
+
"learning_rate": 3.2474226804123714e-05,
|
2304 |
+
"loss": 6.2484,
|
2305 |
+
"step": 656
|
2306 |
+
},
|
2307 |
+
{
|
2308 |
+
"epoch": 6.748717948717949,
|
2309 |
+
"grad_norm": 0.7773774266242981,
|
2310 |
+
"learning_rate": 3.2268041237113405e-05,
|
2311 |
+
"loss": 6.1015,
|
2312 |
+
"step": 658
|
2313 |
+
},
|
2314 |
+
{
|
2315 |
+
"epoch": 6.769230769230769,
|
2316 |
+
"grad_norm": 0.6493405103683472,
|
2317 |
+
"learning_rate": 3.2061855670103095e-05,
|
2318 |
+
"loss": 6.3609,
|
2319 |
+
"step": 660
|
2320 |
+
},
|
2321 |
+
{
|
2322 |
+
"epoch": 6.78974358974359,
|
2323 |
+
"grad_norm": 0.7441650032997131,
|
2324 |
+
"learning_rate": 3.1855670103092785e-05,
|
2325 |
+
"loss": 6.2702,
|
2326 |
+
"step": 662
|
2327 |
+
},
|
2328 |
+
{
|
2329 |
+
"epoch": 6.81025641025641,
|
2330 |
+
"grad_norm": 0.6217365264892578,
|
2331 |
+
"learning_rate": 3.1649484536082475e-05,
|
2332 |
+
"loss": 6.2799,
|
2333 |
+
"step": 664
|
2334 |
+
},
|
2335 |
+
{
|
2336 |
+
"epoch": 6.8307692307692305,
|
2337 |
+
"grad_norm": 0.84025639295578,
|
2338 |
+
"learning_rate": 3.1443298969072166e-05,
|
2339 |
+
"loss": 6.132,
|
2340 |
+
"step": 666
|
2341 |
+
},
|
2342 |
+
{
|
2343 |
+
"epoch": 6.851282051282051,
|
2344 |
+
"grad_norm": 0.7524210214614868,
|
2345 |
+
"learning_rate": 3.1237113402061856e-05,
|
2346 |
+
"loss": 6.1726,
|
2347 |
+
"step": 668
|
2348 |
+
},
|
2349 |
+
{
|
2350 |
+
"epoch": 6.871794871794872,
|
2351 |
+
"grad_norm": 0.6827425956726074,
|
2352 |
+
"learning_rate": 3.1030927835051546e-05,
|
2353 |
+
"loss": 6.2313,
|
2354 |
+
"step": 670
|
2355 |
+
},
|
2356 |
+
{
|
2357 |
+
"epoch": 6.892307692307693,
|
2358 |
+
"grad_norm": 0.6644515991210938,
|
2359 |
+
"learning_rate": 3.082474226804124e-05,
|
2360 |
+
"loss": 6.2285,
|
2361 |
+
"step": 672
|
2362 |
+
},
|
2363 |
+
{
|
2364 |
+
"epoch": 6.912820512820513,
|
2365 |
+
"grad_norm": 0.7661225199699402,
|
2366 |
+
"learning_rate": 3.0618556701030933e-05,
|
2367 |
+
"loss": 6.3277,
|
2368 |
+
"step": 674
|
2369 |
+
},
|
2370 |
+
{
|
2371 |
+
"epoch": 6.933333333333334,
|
2372 |
+
"grad_norm": 0.645756185054779,
|
2373 |
+
"learning_rate": 3.0412371134020617e-05,
|
2374 |
+
"loss": 6.2675,
|
2375 |
+
"step": 676
|
2376 |
+
},
|
2377 |
+
{
|
2378 |
+
"epoch": 6.953846153846154,
|
2379 |
+
"grad_norm": 0.8538247346878052,
|
2380 |
+
"learning_rate": 3.020618556701031e-05,
|
2381 |
+
"loss": 6.0726,
|
2382 |
+
"step": 678
|
2383 |
+
},
|
2384 |
+
{
|
2385 |
+
"epoch": 6.9743589743589745,
|
2386 |
+
"grad_norm": 0.6849417686462402,
|
2387 |
+
"learning_rate": 3e-05,
|
2388 |
+
"loss": 6.2466,
|
2389 |
+
"step": 680
|
2390 |
+
},
|
2391 |
+
{
|
2392 |
+
"epoch": 6.994871794871795,
|
2393 |
+
"grad_norm": 0.7335018515586853,
|
2394 |
+
"learning_rate": 2.979381443298969e-05,
|
2395 |
+
"loss": 6.1795,
|
2396 |
+
"step": 682
|
2397 |
+
},
|
2398 |
+
{
|
2399 |
+
"epoch": 7.015384615384615,
|
2400 |
+
"grad_norm": 0.7281456589698792,
|
2401 |
+
"learning_rate": 2.9587628865979385e-05,
|
2402 |
+
"loss": 6.2103,
|
2403 |
+
"step": 684
|
2404 |
+
},
|
2405 |
+
{
|
2406 |
+
"epoch": 7.035897435897436,
|
2407 |
+
"grad_norm": 0.6370302438735962,
|
2408 |
+
"learning_rate": 2.9381443298969075e-05,
|
2409 |
+
"loss": 6.2624,
|
2410 |
+
"step": 686
|
2411 |
+
},
|
2412 |
+
{
|
2413 |
+
"epoch": 7.056410256410256,
|
2414 |
+
"grad_norm": 0.7385585904121399,
|
2415 |
+
"learning_rate": 2.9175257731958765e-05,
|
2416 |
+
"loss": 6.1058,
|
2417 |
+
"step": 688
|
2418 |
+
},
|
2419 |
+
{
|
2420 |
+
"epoch": 7.076923076923077,
|
2421 |
+
"grad_norm": 0.7029441595077515,
|
2422 |
+
"learning_rate": 2.896907216494846e-05,
|
2423 |
+
"loss": 5.897,
|
2424 |
+
"step": 690
|
2425 |
+
},
|
2426 |
+
{
|
2427 |
+
"epoch": 7.097435897435897,
|
2428 |
+
"grad_norm": 0.7929611802101135,
|
2429 |
+
"learning_rate": 2.8762886597938142e-05,
|
2430 |
+
"loss": 6.0785,
|
2431 |
+
"step": 692
|
2432 |
+
},
|
2433 |
+
{
|
2434 |
+
"epoch": 7.1179487179487175,
|
2435 |
+
"grad_norm": 0.7209746837615967,
|
2436 |
+
"learning_rate": 2.8556701030927836e-05,
|
2437 |
+
"loss": 6.3427,
|
2438 |
+
"step": 694
|
2439 |
+
},
|
2440 |
+
{
|
2441 |
+
"epoch": 7.138461538461539,
|
2442 |
+
"grad_norm": 0.7581877112388611,
|
2443 |
+
"learning_rate": 2.8350515463917526e-05,
|
2444 |
+
"loss": 6.1258,
|
2445 |
+
"step": 696
|
2446 |
+
},
|
2447 |
+
{
|
2448 |
+
"epoch": 7.158974358974359,
|
2449 |
+
"grad_norm": 0.7337011694908142,
|
2450 |
+
"learning_rate": 2.8144329896907216e-05,
|
2451 |
+
"loss": 6.2402,
|
2452 |
+
"step": 698
|
2453 |
+
},
|
2454 |
+
{
|
2455 |
+
"epoch": 7.17948717948718,
|
2456 |
+
"grad_norm": 0.8132451772689819,
|
2457 |
+
"learning_rate": 2.793814432989691e-05,
|
2458 |
+
"loss": 6.2399,
|
2459 |
+
"step": 700
|
2460 |
+
},
|
2461 |
+
{
|
2462 |
+
"epoch": 7.2,
|
2463 |
+
"grad_norm": 0.6459912061691284,
|
2464 |
+
"learning_rate": 2.77319587628866e-05,
|
2465 |
+
"loss": 6.3475,
|
2466 |
+
"step": 702
|
2467 |
+
},
|
2468 |
+
{
|
2469 |
+
"epoch": 7.220512820512821,
|
2470 |
+
"grad_norm": 0.7494829893112183,
|
2471 |
+
"learning_rate": 2.752577319587629e-05,
|
2472 |
+
"loss": 6.1483,
|
2473 |
+
"step": 704
|
2474 |
+
},
|
2475 |
+
{
|
2476 |
+
"epoch": 7.241025641025641,
|
2477 |
+
"grad_norm": 0.7804868221282959,
|
2478 |
+
"learning_rate": 2.7319587628865977e-05,
|
2479 |
+
"loss": 5.9649,
|
2480 |
+
"step": 706
|
2481 |
+
},
|
2482 |
+
{
|
2483 |
+
"epoch": 7.2615384615384615,
|
2484 |
+
"grad_norm": 0.6681525111198425,
|
2485 |
+
"learning_rate": 2.711340206185567e-05,
|
2486 |
+
"loss": 6.1267,
|
2487 |
+
"step": 708
|
2488 |
+
},
|
2489 |
+
{
|
2490 |
+
"epoch": 7.282051282051282,
|
2491 |
+
"grad_norm": 0.8127874732017517,
|
2492 |
+
"learning_rate": 2.690721649484536e-05,
|
2493 |
+
"loss": 6.1324,
|
2494 |
+
"step": 710
|
2495 |
+
},
|
2496 |
+
{
|
2497 |
+
"epoch": 7.302564102564102,
|
2498 |
+
"grad_norm": 0.690651535987854,
|
2499 |
+
"learning_rate": 2.670103092783505e-05,
|
2500 |
+
"loss": 6.3486,
|
2501 |
+
"step": 712
|
2502 |
+
},
|
2503 |
+
{
|
2504 |
+
"epoch": 7.323076923076923,
|
2505 |
+
"grad_norm": 0.7900473475456238,
|
2506 |
+
"learning_rate": 2.6494845360824745e-05,
|
2507 |
+
"loss": 6.2113,
|
2508 |
+
"step": 714
|
2509 |
+
},
|
2510 |
+
{
|
2511 |
+
"epoch": 7.343589743589743,
|
2512 |
+
"grad_norm": 0.9337382316589355,
|
2513 |
+
"learning_rate": 2.6288659793814435e-05,
|
2514 |
+
"loss": 6.2222,
|
2515 |
+
"step": 716
|
2516 |
+
},
|
2517 |
+
{
|
2518 |
+
"epoch": 7.364102564102564,
|
2519 |
+
"grad_norm": 0.6768183708190918,
|
2520 |
+
"learning_rate": 2.6082474226804126e-05,
|
2521 |
+
"loss": 6.0611,
|
2522 |
+
"step": 718
|
2523 |
+
},
|
2524 |
+
{
|
2525 |
+
"epoch": 7.384615384615385,
|
2526 |
+
"grad_norm": 0.8045716881752014,
|
2527 |
+
"learning_rate": 2.587628865979382e-05,
|
2528 |
+
"loss": 6.1869,
|
2529 |
+
"step": 720
|
2530 |
+
},
|
2531 |
+
{
|
2532 |
+
"epoch": 7.4051282051282055,
|
2533 |
+
"grad_norm": 0.7803860902786255,
|
2534 |
+
"learning_rate": 2.5670103092783503e-05,
|
2535 |
+
"loss": 6.2456,
|
2536 |
+
"step": 722
|
2537 |
+
},
|
2538 |
+
{
|
2539 |
+
"epoch": 7.425641025641026,
|
2540 |
+
"grad_norm": 0.6583536863327026,
|
2541 |
+
"learning_rate": 2.5463917525773196e-05,
|
2542 |
+
"loss": 6.3823,
|
2543 |
+
"step": 724
|
2544 |
+
},
|
2545 |
+
{
|
2546 |
+
"epoch": 7.446153846153846,
|
2547 |
+
"grad_norm": 0.6829738616943359,
|
2548 |
+
"learning_rate": 2.5257731958762887e-05,
|
2549 |
+
"loss": 5.9673,
|
2550 |
+
"step": 726
|
2551 |
+
},
|
2552 |
+
{
|
2553 |
+
"epoch": 7.466666666666667,
|
2554 |
+
"grad_norm": 0.6171258091926575,
|
2555 |
+
"learning_rate": 2.5051546391752577e-05,
|
2556 |
+
"loss": 6.4778,
|
2557 |
+
"step": 728
|
2558 |
+
},
|
2559 |
+
{
|
2560 |
+
"epoch": 7.487179487179487,
|
2561 |
+
"grad_norm": 0.9100000858306885,
|
2562 |
+
"learning_rate": 2.484536082474227e-05,
|
2563 |
+
"loss": 6.1923,
|
2564 |
+
"step": 730
|
2565 |
+
},
|
2566 |
+
{
|
2567 |
+
"epoch": 7.507692307692308,
|
2568 |
+
"grad_norm": 0.6728296279907227,
|
2569 |
+
"learning_rate": 2.463917525773196e-05,
|
2570 |
+
"loss": 6.1936,
|
2571 |
+
"step": 732
|
2572 |
+
},
|
2573 |
+
{
|
2574 |
+
"epoch": 7.528205128205128,
|
2575 |
+
"grad_norm": 0.6537327766418457,
|
2576 |
+
"learning_rate": 2.443298969072165e-05,
|
2577 |
+
"loss": 6.3321,
|
2578 |
+
"step": 734
|
2579 |
+
},
|
2580 |
+
{
|
2581 |
+
"epoch": 7.5487179487179485,
|
2582 |
+
"grad_norm": 0.7532163858413696,
|
2583 |
+
"learning_rate": 2.422680412371134e-05,
|
2584 |
+
"loss": 6.263,
|
2585 |
+
"step": 736
|
2586 |
+
},
|
2587 |
+
{
|
2588 |
+
"epoch": 7.569230769230769,
|
2589 |
+
"grad_norm": 0.8065058588981628,
|
2590 |
+
"learning_rate": 2.402061855670103e-05,
|
2591 |
+
"loss": 6.187,
|
2592 |
+
"step": 738
|
2593 |
+
},
|
2594 |
+
{
|
2595 |
+
"epoch": 7.589743589743589,
|
2596 |
+
"grad_norm": 0.6982044577598572,
|
2597 |
+
"learning_rate": 2.3814432989690722e-05,
|
2598 |
+
"loss": 6.2071,
|
2599 |
+
"step": 740
|
2600 |
+
},
|
2601 |
+
{
|
2602 |
+
"epoch": 7.61025641025641,
|
2603 |
+
"grad_norm": 0.7518334984779358,
|
2604 |
+
"learning_rate": 2.3608247422680412e-05,
|
2605 |
+
"loss": 6.1698,
|
2606 |
+
"step": 742
|
2607 |
+
},
|
2608 |
+
{
|
2609 |
+
"epoch": 7.63076923076923,
|
2610 |
+
"grad_norm": 0.6980510950088501,
|
2611 |
+
"learning_rate": 2.3402061855670106e-05,
|
2612 |
+
"loss": 6.1269,
|
2613 |
+
"step": 744
|
2614 |
+
},
|
2615 |
+
{
|
2616 |
+
"epoch": 7.651282051282052,
|
2617 |
+
"grad_norm": 0.7211715579032898,
|
2618 |
+
"learning_rate": 2.3195876288659796e-05,
|
2619 |
+
"loss": 6.2005,
|
2620 |
+
"step": 746
|
2621 |
+
},
|
2622 |
+
{
|
2623 |
+
"epoch": 7.671794871794872,
|
2624 |
+
"grad_norm": 0.7351675033569336,
|
2625 |
+
"learning_rate": 2.2989690721649483e-05,
|
2626 |
+
"loss": 6.1096,
|
2627 |
+
"step": 748
|
2628 |
+
},
|
2629 |
+
{
|
2630 |
+
"epoch": 7.6923076923076925,
|
2631 |
+
"grad_norm": 0.6142987608909607,
|
2632 |
+
"learning_rate": 2.2783505154639176e-05,
|
2633 |
+
"loss": 6.0628,
|
2634 |
+
"step": 750
|
2635 |
+
},
|
2636 |
+
{
|
2637 |
+
"epoch": 7.712820512820513,
|
2638 |
+
"grad_norm": 0.6700417995452881,
|
2639 |
+
"learning_rate": 2.2577319587628867e-05,
|
2640 |
+
"loss": 6.2528,
|
2641 |
+
"step": 752
|
2642 |
+
},
|
2643 |
+
{
|
2644 |
+
"epoch": 7.733333333333333,
|
2645 |
+
"grad_norm": 0.7297637462615967,
|
2646 |
+
"learning_rate": 2.237113402061856e-05,
|
2647 |
+
"loss": 6.165,
|
2648 |
+
"step": 754
|
2649 |
+
},
|
2650 |
+
{
|
2651 |
+
"epoch": 7.753846153846154,
|
2652 |
+
"grad_norm": 0.6762688755989075,
|
2653 |
+
"learning_rate": 2.2164948453608247e-05,
|
2654 |
+
"loss": 6.3651,
|
2655 |
+
"step": 756
|
2656 |
+
},
|
2657 |
+
{
|
2658 |
+
"epoch": 7.774358974358974,
|
2659 |
+
"grad_norm": 0.6471550464630127,
|
2660 |
+
"learning_rate": 2.1958762886597937e-05,
|
2661 |
+
"loss": 6.142,
|
2662 |
+
"step": 758
|
2663 |
+
},
|
2664 |
+
{
|
2665 |
+
"epoch": 7.794871794871795,
|
2666 |
+
"grad_norm": 0.6905655264854431,
|
2667 |
+
"learning_rate": 2.175257731958763e-05,
|
2668 |
+
"loss": 6.4471,
|
2669 |
+
"step": 760
|
2670 |
+
},
|
2671 |
+
{
|
2672 |
+
"epoch": 7.815384615384615,
|
2673 |
+
"grad_norm": 0.6673563122749329,
|
2674 |
+
"learning_rate": 2.154639175257732e-05,
|
2675 |
+
"loss": 6.0799,
|
2676 |
+
"step": 762
|
2677 |
+
},
|
2678 |
+
{
|
2679 |
+
"epoch": 7.835897435897436,
|
2680 |
+
"grad_norm": 0.7106631994247437,
|
2681 |
+
"learning_rate": 2.134020618556701e-05,
|
2682 |
+
"loss": 6.3385,
|
2683 |
+
"step": 764
|
2684 |
+
},
|
2685 |
+
{
|
2686 |
+
"epoch": 7.856410256410256,
|
2687 |
+
"grad_norm": 0.7296525239944458,
|
2688 |
+
"learning_rate": 2.1134020618556702e-05,
|
2689 |
+
"loss": 6.1772,
|
2690 |
+
"step": 766
|
2691 |
+
},
|
2692 |
+
{
|
2693 |
+
"epoch": 7.876923076923077,
|
2694 |
+
"grad_norm": 0.6923534274101257,
|
2695 |
+
"learning_rate": 2.0927835051546392e-05,
|
2696 |
+
"loss": 6.0845,
|
2697 |
+
"step": 768
|
2698 |
+
},
|
2699 |
+
{
|
2700 |
+
"epoch": 7.897435897435898,
|
2701 |
+
"grad_norm": 0.6479873061180115,
|
2702 |
+
"learning_rate": 2.0721649484536086e-05,
|
2703 |
+
"loss": 6.4237,
|
2704 |
+
"step": 770
|
2705 |
+
},
|
2706 |
+
{
|
2707 |
+
"epoch": 7.917948717948718,
|
2708 |
+
"grad_norm": 0.7043232321739197,
|
2709 |
+
"learning_rate": 2.0515463917525773e-05,
|
2710 |
+
"loss": 6.2389,
|
2711 |
+
"step": 772
|
2712 |
+
},
|
2713 |
+
{
|
2714 |
+
"epoch": 7.938461538461539,
|
2715 |
+
"grad_norm": 0.6955374479293823,
|
2716 |
+
"learning_rate": 2.0309278350515466e-05,
|
2717 |
+
"loss": 6.2594,
|
2718 |
+
"step": 774
|
2719 |
+
},
|
2720 |
+
{
|
2721 |
+
"epoch": 7.958974358974359,
|
2722 |
+
"grad_norm": 0.6248512864112854,
|
2723 |
+
"learning_rate": 2.0103092783505157e-05,
|
2724 |
+
"loss": 6.3424,
|
2725 |
+
"step": 776
|
2726 |
+
},
|
2727 |
+
{
|
2728 |
+
"epoch": 7.97948717948718,
|
2729 |
+
"grad_norm": 0.7696142196655273,
|
2730 |
+
"learning_rate": 1.9896907216494843e-05,
|
2731 |
+
"loss": 6.0404,
|
2732 |
+
"step": 778
|
2733 |
+
},
|
2734 |
+
{
|
2735 |
+
"epoch": 8.0,
|
2736 |
+
"grad_norm": 0.7960737347602844,
|
2737 |
+
"learning_rate": 1.9690721649484537e-05,
|
2738 |
+
"loss": 6.2037,
|
2739 |
+
"step": 780
|
2740 |
+
},
|
2741 |
+
{
|
2742 |
+
"epoch": 8.02051282051282,
|
2743 |
+
"grad_norm": 0.764180600643158,
|
2744 |
+
"learning_rate": 1.9484536082474227e-05,
|
2745 |
+
"loss": 6.0193,
|
2746 |
+
"step": 782
|
2747 |
+
},
|
2748 |
+
{
|
2749 |
+
"epoch": 8.04102564102564,
|
2750 |
+
"grad_norm": 0.7745684385299683,
|
2751 |
+
"learning_rate": 1.927835051546392e-05,
|
2752 |
+
"loss": 6.1723,
|
2753 |
+
"step": 784
|
2754 |
+
},
|
2755 |
+
{
|
2756 |
+
"epoch": 8.061538461538461,
|
2757 |
+
"grad_norm": 0.7468089461326599,
|
2758 |
+
"learning_rate": 1.9072164948453608e-05,
|
2759 |
+
"loss": 5.989,
|
2760 |
+
"step": 786
|
2761 |
+
},
|
2762 |
+
{
|
2763 |
+
"epoch": 8.082051282051282,
|
2764 |
+
"grad_norm": 0.7076855301856995,
|
2765 |
+
"learning_rate": 1.8865979381443298e-05,
|
2766 |
+
"loss": 6.2975,
|
2767 |
+
"step": 788
|
2768 |
+
},
|
2769 |
+
{
|
2770 |
+
"epoch": 8.102564102564102,
|
2771 |
+
"grad_norm": 0.750068187713623,
|
2772 |
+
"learning_rate": 1.865979381443299e-05,
|
2773 |
+
"loss": 6.2057,
|
2774 |
+
"step": 790
|
2775 |
+
},
|
2776 |
+
{
|
2777 |
+
"epoch": 8.123076923076923,
|
2778 |
+
"grad_norm": 0.5782350301742554,
|
2779 |
+
"learning_rate": 1.8453608247422682e-05,
|
2780 |
+
"loss": 6.3482,
|
2781 |
+
"step": 792
|
2782 |
+
},
|
2783 |
+
{
|
2784 |
+
"epoch": 8.143589743589743,
|
2785 |
+
"grad_norm": 0.7477689981460571,
|
2786 |
+
"learning_rate": 1.8247422680412372e-05,
|
2787 |
+
"loss": 6.1132,
|
2788 |
+
"step": 794
|
2789 |
+
},
|
2790 |
+
{
|
2791 |
+
"epoch": 8.164102564102564,
|
2792 |
+
"grad_norm": 0.7583786845207214,
|
2793 |
+
"learning_rate": 1.8041237113402062e-05,
|
2794 |
+
"loss": 6.3927,
|
2795 |
+
"step": 796
|
2796 |
+
},
|
2797 |
+
{
|
2798 |
+
"epoch": 8.184615384615384,
|
2799 |
+
"grad_norm": 1.066887378692627,
|
2800 |
+
"learning_rate": 1.7835051546391753e-05,
|
2801 |
+
"loss": 6.3036,
|
2802 |
+
"step": 798
|
2803 |
+
},
|
2804 |
+
{
|
2805 |
+
"epoch": 8.205128205128204,
|
2806 |
+
"grad_norm": 0.7979178428649902,
|
2807 |
+
"learning_rate": 1.7628865979381446e-05,
|
2808 |
+
"loss": 6.0357,
|
2809 |
+
"step": 800
|
2810 |
+
},
|
2811 |
+
{
|
2812 |
+
"epoch": 8.225641025641025,
|
2813 |
+
"grad_norm": 0.7925027012825012,
|
2814 |
+
"learning_rate": 1.7422680412371133e-05,
|
2815 |
+
"loss": 6.1571,
|
2816 |
+
"step": 802
|
2817 |
+
},
|
2818 |
+
{
|
2819 |
+
"epoch": 8.246153846153845,
|
2820 |
+
"grad_norm": 0.6301718950271606,
|
2821 |
+
"learning_rate": 1.7216494845360827e-05,
|
2822 |
+
"loss": 6.2834,
|
2823 |
+
"step": 804
|
2824 |
+
},
|
2825 |
+
{
|
2826 |
+
"epoch": 8.266666666666667,
|
2827 |
+
"grad_norm": 0.8579639196395874,
|
2828 |
+
"learning_rate": 1.7010309278350517e-05,
|
2829 |
+
"loss": 6.1507,
|
2830 |
+
"step": 806
|
2831 |
+
},
|
2832 |
+
{
|
2833 |
+
"epoch": 8.287179487179488,
|
2834 |
+
"grad_norm": 0.6312888264656067,
|
2835 |
+
"learning_rate": 1.6804123711340207e-05,
|
2836 |
+
"loss": 6.4582,
|
2837 |
+
"step": 808
|
2838 |
+
},
|
2839 |
+
{
|
2840 |
+
"epoch": 8.307692307692308,
|
2841 |
+
"grad_norm": 0.7432986497879028,
|
2842 |
+
"learning_rate": 1.6597938144329898e-05,
|
2843 |
+
"loss": 6.2031,
|
2844 |
+
"step": 810
|
2845 |
+
},
|
2846 |
+
{
|
2847 |
+
"epoch": 8.328205128205129,
|
2848 |
+
"grad_norm": 0.6533825993537903,
|
2849 |
+
"learning_rate": 1.6391752577319588e-05,
|
2850 |
+
"loss": 6.2383,
|
2851 |
+
"step": 812
|
2852 |
+
},
|
2853 |
+
{
|
2854 |
+
"epoch": 8.34871794871795,
|
2855 |
+
"grad_norm": 0.7833907604217529,
|
2856 |
+
"learning_rate": 1.618556701030928e-05,
|
2857 |
+
"loss": 6.2307,
|
2858 |
+
"step": 814
|
2859 |
+
},
|
2860 |
+
{
|
2861 |
+
"epoch": 8.36923076923077,
|
2862 |
+
"grad_norm": 0.7582297325134277,
|
2863 |
+
"learning_rate": 1.5979381443298968e-05,
|
2864 |
+
"loss": 6.2063,
|
2865 |
+
"step": 816
|
2866 |
+
},
|
2867 |
+
{
|
2868 |
+
"epoch": 8.38974358974359,
|
2869 |
+
"grad_norm": 0.7262023687362671,
|
2870 |
+
"learning_rate": 1.577319587628866e-05,
|
2871 |
+
"loss": 6.2125,
|
2872 |
+
"step": 818
|
2873 |
+
},
|
2874 |
+
{
|
2875 |
+
"epoch": 8.41025641025641,
|
2876 |
+
"grad_norm": 0.833846926689148,
|
2877 |
+
"learning_rate": 1.5567010309278352e-05,
|
2878 |
+
"loss": 5.9073,
|
2879 |
+
"step": 820
|
2880 |
+
},
|
2881 |
+
{
|
2882 |
+
"epoch": 8.430769230769231,
|
2883 |
+
"grad_norm": 0.6528037190437317,
|
2884 |
+
"learning_rate": 1.5360824742268042e-05,
|
2885 |
+
"loss": 6.1687,
|
2886 |
+
"step": 822
|
2887 |
+
},
|
2888 |
+
{
|
2889 |
+
"epoch": 8.451282051282051,
|
2890 |
+
"grad_norm": 0.715062141418457,
|
2891 |
+
"learning_rate": 1.5154639175257731e-05,
|
2892 |
+
"loss": 6.3858,
|
2893 |
+
"step": 824
|
2894 |
+
},
|
2895 |
+
{
|
2896 |
+
"epoch": 8.471794871794872,
|
2897 |
+
"grad_norm": 0.6965751647949219,
|
2898 |
+
"learning_rate": 1.4948453608247423e-05,
|
2899 |
+
"loss": 6.2247,
|
2900 |
+
"step": 826
|
2901 |
+
},
|
2902 |
+
{
|
2903 |
+
"epoch": 8.492307692307692,
|
2904 |
+
"grad_norm": 0.644185483455658,
|
2905 |
+
"learning_rate": 1.4742268041237115e-05,
|
2906 |
+
"loss": 6.4232,
|
2907 |
+
"step": 828
|
2908 |
+
},
|
2909 |
+
{
|
2910 |
+
"epoch": 8.512820512820513,
|
2911 |
+
"grad_norm": 0.8024539351463318,
|
2912 |
+
"learning_rate": 1.4536082474226805e-05,
|
2913 |
+
"loss": 6.2306,
|
2914 |
+
"step": 830
|
2915 |
+
},
|
2916 |
+
{
|
2917 |
+
"epoch": 8.533333333333333,
|
2918 |
+
"grad_norm": 0.6547759175300598,
|
2919 |
+
"learning_rate": 1.4329896907216495e-05,
|
2920 |
+
"loss": 6.2272,
|
2921 |
+
"step": 832
|
2922 |
+
},
|
2923 |
+
{
|
2924 |
+
"epoch": 8.553846153846154,
|
2925 |
+
"grad_norm": 0.8308131694793701,
|
2926 |
+
"learning_rate": 1.4123711340206186e-05,
|
2927 |
+
"loss": 6.2751,
|
2928 |
+
"step": 834
|
2929 |
+
},
|
2930 |
+
{
|
2931 |
+
"epoch": 8.574358974358974,
|
2932 |
+
"grad_norm": 0.6475526094436646,
|
2933 |
+
"learning_rate": 1.3917525773195878e-05,
|
2934 |
+
"loss": 6.3477,
|
2935 |
+
"step": 836
|
2936 |
+
},
|
2937 |
+
{
|
2938 |
+
"epoch": 8.594871794871795,
|
2939 |
+
"grad_norm": 0.7138957977294922,
|
2940 |
+
"learning_rate": 1.371134020618557e-05,
|
2941 |
+
"loss": 6.2169,
|
2942 |
+
"step": 838
|
2943 |
+
},
|
2944 |
+
{
|
2945 |
+
"epoch": 8.615384615384615,
|
2946 |
+
"grad_norm": 0.7940370440483093,
|
2947 |
+
"learning_rate": 1.3505154639175258e-05,
|
2948 |
+
"loss": 6.0711,
|
2949 |
+
"step": 840
|
2950 |
+
},
|
2951 |
+
{
|
2952 |
+
"epoch": 8.635897435897435,
|
2953 |
+
"grad_norm": 0.710660457611084,
|
2954 |
+
"learning_rate": 1.3298969072164948e-05,
|
2955 |
+
"loss": 6.1328,
|
2956 |
+
"step": 842
|
2957 |
+
},
|
2958 |
+
{
|
2959 |
+
"epoch": 8.656410256410256,
|
2960 |
+
"grad_norm": 0.6740992069244385,
|
2961 |
+
"learning_rate": 1.309278350515464e-05,
|
2962 |
+
"loss": 6.2399,
|
2963 |
+
"step": 844
|
2964 |
+
},
|
2965 |
+
{
|
2966 |
+
"epoch": 8.676923076923076,
|
2967 |
+
"grad_norm": 0.6954624652862549,
|
2968 |
+
"learning_rate": 1.2886597938144329e-05,
|
2969 |
+
"loss": 5.8876,
|
2970 |
+
"step": 846
|
2971 |
+
},
|
2972 |
+
{
|
2973 |
+
"epoch": 8.697435897435897,
|
2974 |
+
"grad_norm": 0.6282210350036621,
|
2975 |
+
"learning_rate": 1.268041237113402e-05,
|
2976 |
+
"loss": 6.2823,
|
2977 |
+
"step": 848
|
2978 |
+
},
|
2979 |
+
{
|
2980 |
+
"epoch": 8.717948717948717,
|
2981 |
+
"grad_norm": 0.6641991138458252,
|
2982 |
+
"learning_rate": 1.2474226804123713e-05,
|
2983 |
+
"loss": 6.0473,
|
2984 |
+
"step": 850
|
2985 |
+
},
|
2986 |
+
{
|
2987 |
+
"epoch": 8.73846153846154,
|
2988 |
+
"grad_norm": 0.9313386678695679,
|
2989 |
+
"learning_rate": 1.2268041237113401e-05,
|
2990 |
+
"loss": 5.9459,
|
2991 |
+
"step": 852
|
2992 |
+
},
|
2993 |
+
{
|
2994 |
+
"epoch": 8.75897435897436,
|
2995 |
+
"grad_norm": 0.6434590816497803,
|
2996 |
+
"learning_rate": 1.2061855670103093e-05,
|
2997 |
+
"loss": 6.2093,
|
2998 |
+
"step": 854
|
2999 |
+
},
|
3000 |
+
{
|
3001 |
+
"epoch": 8.77948717948718,
|
3002 |
+
"grad_norm": 0.7539594173431396,
|
3003 |
+
"learning_rate": 1.1855670103092783e-05,
|
3004 |
+
"loss": 6.1687,
|
3005 |
+
"step": 856
|
3006 |
+
},
|
3007 |
+
{
|
3008 |
+
"epoch": 8.8,
|
3009 |
+
"grad_norm": 0.7946255803108215,
|
3010 |
+
"learning_rate": 1.1649484536082475e-05,
|
3011 |
+
"loss": 6.1376,
|
3012 |
+
"step": 858
|
3013 |
+
},
|
3014 |
+
{
|
3015 |
+
"epoch": 8.820512820512821,
|
3016 |
+
"grad_norm": 0.7640373706817627,
|
3017 |
+
"learning_rate": 1.1443298969072166e-05,
|
3018 |
+
"loss": 6.1806,
|
3019 |
+
"step": 860
|
3020 |
+
},
|
3021 |
+
{
|
3022 |
+
"epoch": 8.841025641025642,
|
3023 |
+
"grad_norm": 0.6656478643417358,
|
3024 |
+
"learning_rate": 1.1237113402061856e-05,
|
3025 |
+
"loss": 6.0972,
|
3026 |
+
"step": 862
|
3027 |
+
},
|
3028 |
+
{
|
3029 |
+
"epoch": 8.861538461538462,
|
3030 |
+
"grad_norm": 0.7396649718284607,
|
3031 |
+
"learning_rate": 1.1030927835051546e-05,
|
3032 |
+
"loss": 6.0908,
|
3033 |
+
"step": 864
|
3034 |
+
},
|
3035 |
+
{
|
3036 |
+
"epoch": 8.882051282051282,
|
3037 |
+
"grad_norm": 0.6619882583618164,
|
3038 |
+
"learning_rate": 1.0824742268041238e-05,
|
3039 |
+
"loss": 6.4136,
|
3040 |
+
"step": 866
|
3041 |
+
},
|
3042 |
+
{
|
3043 |
+
"epoch": 8.902564102564103,
|
3044 |
+
"grad_norm": 0.7462313175201416,
|
3045 |
+
"learning_rate": 1.0618556701030928e-05,
|
3046 |
+
"loss": 6.0462,
|
3047 |
+
"step": 868
|
3048 |
+
},
|
3049 |
+
{
|
3050 |
+
"epoch": 8.923076923076923,
|
3051 |
+
"grad_norm": 1.2745590209960938,
|
3052 |
+
"learning_rate": 1.041237113402062e-05,
|
3053 |
+
"loss": 6.0757,
|
3054 |
+
"step": 870
|
3055 |
+
},
|
3056 |
+
{
|
3057 |
+
"epoch": 8.943589743589744,
|
3058 |
+
"grad_norm": 0.7905880212783813,
|
3059 |
+
"learning_rate": 1.0206185567010309e-05,
|
3060 |
+
"loss": 5.9552,
|
3061 |
+
"step": 872
|
3062 |
+
},
|
3063 |
+
{
|
3064 |
+
"epoch": 8.964102564102564,
|
3065 |
+
"grad_norm": 0.7253865003585815,
|
3066 |
+
"learning_rate": 1e-05,
|
3067 |
+
"loss": 6.2264,
|
3068 |
+
"step": 874
|
3069 |
+
},
|
3070 |
+
{
|
3071 |
+
"epoch": 8.984615384615385,
|
3072 |
+
"grad_norm": 0.7785334587097168,
|
3073 |
+
"learning_rate": 9.793814432989691e-06,
|
3074 |
+
"loss": 6.195,
|
3075 |
+
"step": 876
|
3076 |
+
},
|
3077 |
+
{
|
3078 |
+
"epoch": 9.005128205128205,
|
3079 |
+
"grad_norm": 0.8180606961250305,
|
3080 |
+
"learning_rate": 9.587628865979383e-06,
|
3081 |
+
"loss": 5.991,
|
3082 |
+
"step": 878
|
3083 |
+
},
|
3084 |
+
{
|
3085 |
+
"epoch": 9.025641025641026,
|
3086 |
+
"grad_norm": 0.7430230379104614,
|
3087 |
+
"learning_rate": 9.381443298969073e-06,
|
3088 |
+
"loss": 5.9541,
|
3089 |
+
"step": 880
|
3090 |
+
},
|
3091 |
+
{
|
3092 |
+
"epoch": 9.046153846153846,
|
3093 |
+
"grad_norm": 0.6766589879989624,
|
3094 |
+
"learning_rate": 9.175257731958762e-06,
|
3095 |
+
"loss": 6.2283,
|
3096 |
+
"step": 882
|
3097 |
+
},
|
3098 |
+
{
|
3099 |
+
"epoch": 9.066666666666666,
|
3100 |
+
"grad_norm": 0.7492343187332153,
|
3101 |
+
"learning_rate": 8.969072164948454e-06,
|
3102 |
+
"loss": 6.1433,
|
3103 |
+
"step": 884
|
3104 |
+
},
|
3105 |
+
{
|
3106 |
+
"epoch": 9.087179487179487,
|
3107 |
+
"grad_norm": 0.787506103515625,
|
3108 |
+
"learning_rate": 8.762886597938144e-06,
|
3109 |
+
"loss": 5.7677,
|
3110 |
+
"step": 886
|
3111 |
+
},
|
3112 |
+
{
|
3113 |
+
"epoch": 9.107692307692307,
|
3114 |
+
"grad_norm": 0.6319190859794617,
|
3115 |
+
"learning_rate": 8.556701030927836e-06,
|
3116 |
+
"loss": 6.2569,
|
3117 |
+
"step": 888
|
3118 |
+
},
|
3119 |
+
{
|
3120 |
+
"epoch": 9.128205128205128,
|
3121 |
+
"grad_norm": 0.5653165578842163,
|
3122 |
+
"learning_rate": 8.350515463917526e-06,
|
3123 |
+
"loss": 6.3534,
|
3124 |
+
"step": 890
|
3125 |
+
},
|
3126 |
+
{
|
3127 |
+
"epoch": 9.148717948717948,
|
3128 |
+
"grad_norm": 1.010986566543579,
|
3129 |
+
"learning_rate": 8.144329896907216e-06,
|
3130 |
+
"loss": 6.2486,
|
3131 |
+
"step": 892
|
3132 |
+
},
|
3133 |
+
{
|
3134 |
+
"epoch": 9.169230769230769,
|
3135 |
+
"grad_norm": 0.7131024599075317,
|
3136 |
+
"learning_rate": 7.938144329896907e-06,
|
3137 |
+
"loss": 6.0887,
|
3138 |
+
"step": 894
|
3139 |
+
},
|
3140 |
+
{
|
3141 |
+
"epoch": 9.189743589743589,
|
3142 |
+
"grad_norm": 0.7308672666549683,
|
3143 |
+
"learning_rate": 7.731958762886599e-06,
|
3144 |
+
"loss": 6.1484,
|
3145 |
+
"step": 896
|
3146 |
+
},
|
3147 |
+
{
|
3148 |
+
"epoch": 9.21025641025641,
|
3149 |
+
"grad_norm": 0.7401643991470337,
|
3150 |
+
"learning_rate": 7.525773195876289e-06,
|
3151 |
+
"loss": 6.1046,
|
3152 |
+
"step": 898
|
3153 |
+
},
|
3154 |
+
{
|
3155 |
+
"epoch": 9.23076923076923,
|
3156 |
+
"grad_norm": 0.7252262830734253,
|
3157 |
+
"learning_rate": 7.31958762886598e-06,
|
3158 |
+
"loss": 6.2832,
|
3159 |
+
"step": 900
|
3160 |
+
},
|
3161 |
+
{
|
3162 |
+
"epoch": 9.25128205128205,
|
3163 |
+
"grad_norm": 0.7304520606994629,
|
3164 |
+
"learning_rate": 7.11340206185567e-06,
|
3165 |
+
"loss": 6.2027,
|
3166 |
+
"step": 902
|
3167 |
+
},
|
3168 |
+
{
|
3169 |
+
"epoch": 9.271794871794873,
|
3170 |
+
"grad_norm": 0.6221430897712708,
|
3171 |
+
"learning_rate": 6.907216494845361e-06,
|
3172 |
+
"loss": 6.3802,
|
3173 |
+
"step": 904
|
3174 |
+
},
|
3175 |
+
{
|
3176 |
+
"epoch": 9.292307692307693,
|
3177 |
+
"grad_norm": 0.7544053792953491,
|
3178 |
+
"learning_rate": 6.701030927835052e-06,
|
3179 |
+
"loss": 6.1048,
|
3180 |
+
"step": 906
|
3181 |
+
},
|
3182 |
+
{
|
3183 |
+
"epoch": 9.312820512820513,
|
3184 |
+
"grad_norm": 0.6470032930374146,
|
3185 |
+
"learning_rate": 6.494845360824743e-06,
|
3186 |
+
"loss": 6.3017,
|
3187 |
+
"step": 908
|
3188 |
+
},
|
3189 |
+
{
|
3190 |
+
"epoch": 9.333333333333334,
|
3191 |
+
"grad_norm": 0.5889255404472351,
|
3192 |
+
"learning_rate": 6.288659793814433e-06,
|
3193 |
+
"loss": 6.2106,
|
3194 |
+
"step": 910
|
3195 |
+
},
|
3196 |
+
{
|
3197 |
+
"epoch": 9.353846153846154,
|
3198 |
+
"grad_norm": 0.6223869323730469,
|
3199 |
+
"learning_rate": 6.082474226804124e-06,
|
3200 |
+
"loss": 6.1163,
|
3201 |
+
"step": 912
|
3202 |
+
},
|
3203 |
+
{
|
3204 |
+
"epoch": 9.374358974358975,
|
3205 |
+
"grad_norm": 0.6783230304718018,
|
3206 |
+
"learning_rate": 5.876288659793814e-06,
|
3207 |
+
"loss": 6.1365,
|
3208 |
+
"step": 914
|
3209 |
+
},
|
3210 |
+
{
|
3211 |
+
"epoch": 9.394871794871795,
|
3212 |
+
"grad_norm": 0.7757629752159119,
|
3213 |
+
"learning_rate": 5.670103092783505e-06,
|
3214 |
+
"loss": 6.0445,
|
3215 |
+
"step": 916
|
3216 |
+
},
|
3217 |
+
{
|
3218 |
+
"epoch": 9.415384615384616,
|
3219 |
+
"grad_norm": 0.7045403122901917,
|
3220 |
+
"learning_rate": 5.4639175257731965e-06,
|
3221 |
+
"loss": 6.3132,
|
3222 |
+
"step": 918
|
3223 |
+
},
|
3224 |
+
{
|
3225 |
+
"epoch": 9.435897435897436,
|
3226 |
+
"grad_norm": 0.7770035862922668,
|
3227 |
+
"learning_rate": 5.257731958762887e-06,
|
3228 |
+
"loss": 6.0804,
|
3229 |
+
"step": 920
|
3230 |
+
},
|
3231 |
+
{
|
3232 |
+
"epoch": 9.456410256410257,
|
3233 |
+
"grad_norm": 0.6999940276145935,
|
3234 |
+
"learning_rate": 5.051546391752578e-06,
|
3235 |
+
"loss": 6.0772,
|
3236 |
+
"step": 922
|
3237 |
+
},
|
3238 |
+
{
|
3239 |
+
"epoch": 9.476923076923077,
|
3240 |
+
"grad_norm": 0.7113653421401978,
|
3241 |
+
"learning_rate": 4.845360824742268e-06,
|
3242 |
+
"loss": 5.9905,
|
3243 |
+
"step": 924
|
3244 |
+
},
|
3245 |
+
{
|
3246 |
+
"epoch": 9.497435897435897,
|
3247 |
+
"grad_norm": 0.6142096519470215,
|
3248 |
+
"learning_rate": 4.639175257731959e-06,
|
3249 |
+
"loss": 6.2268,
|
3250 |
+
"step": 926
|
3251 |
+
},
|
3252 |
+
{
|
3253 |
+
"epoch": 9.517948717948718,
|
3254 |
+
"grad_norm": 0.6630533933639526,
|
3255 |
+
"learning_rate": 4.43298969072165e-06,
|
3256 |
+
"loss": 6.3161,
|
3257 |
+
"step": 928
|
3258 |
+
},
|
3259 |
+
{
|
3260 |
+
"epoch": 9.538461538461538,
|
3261 |
+
"grad_norm": 0.8330481052398682,
|
3262 |
+
"learning_rate": 4.2268041237113405e-06,
|
3263 |
+
"loss": 6.1675,
|
3264 |
+
"step": 930
|
3265 |
+
},
|
3266 |
+
{
|
3267 |
+
"epoch": 9.558974358974359,
|
3268 |
+
"grad_norm": 0.6822926998138428,
|
3269 |
+
"learning_rate": 4.020618556701032e-06,
|
3270 |
+
"loss": 6.106,
|
3271 |
+
"step": 932
|
3272 |
+
},
|
3273 |
+
{
|
3274 |
+
"epoch": 9.57948717948718,
|
3275 |
+
"grad_norm": 0.7913044095039368,
|
3276 |
+
"learning_rate": 3.8144329896907223e-06,
|
3277 |
+
"loss": 6.2236,
|
3278 |
+
"step": 934
|
3279 |
+
},
|
3280 |
+
{
|
3281 |
+
"epoch": 9.6,
|
3282 |
+
"grad_norm": 0.8586515188217163,
|
3283 |
+
"learning_rate": 3.608247422680412e-06,
|
3284 |
+
"loss": 6.0912,
|
3285 |
+
"step": 936
|
3286 |
+
},
|
3287 |
+
{
|
3288 |
+
"epoch": 9.62051282051282,
|
3289 |
+
"grad_norm": 0.6389515399932861,
|
3290 |
+
"learning_rate": 3.402061855670103e-06,
|
3291 |
+
"loss": 6.1972,
|
3292 |
+
"step": 938
|
3293 |
+
},
|
3294 |
+
{
|
3295 |
+
"epoch": 9.64102564102564,
|
3296 |
+
"grad_norm": 0.647665798664093,
|
3297 |
+
"learning_rate": 3.195876288659794e-06,
|
3298 |
+
"loss": 6.462,
|
3299 |
+
"step": 940
|
3300 |
+
},
|
3301 |
+
{
|
3302 |
+
"epoch": 9.661538461538461,
|
3303 |
+
"grad_norm": 0.7906177043914795,
|
3304 |
+
"learning_rate": 2.989690721649485e-06,
|
3305 |
+
"loss": 6.1017,
|
3306 |
+
"step": 942
|
3307 |
+
},
|
3308 |
+
{
|
3309 |
+
"epoch": 9.682051282051281,
|
3310 |
+
"grad_norm": 0.7301990985870361,
|
3311 |
+
"learning_rate": 2.7835051546391757e-06,
|
3312 |
+
"loss": 6.1658,
|
3313 |
+
"step": 944
|
3314 |
+
},
|
3315 |
+
{
|
3316 |
+
"epoch": 9.702564102564102,
|
3317 |
+
"grad_norm": 0.676216185092926,
|
3318 |
+
"learning_rate": 2.577319587628866e-06,
|
3319 |
+
"loss": 6.3423,
|
3320 |
+
"step": 946
|
3321 |
+
},
|
3322 |
+
{
|
3323 |
+
"epoch": 9.723076923076922,
|
3324 |
+
"grad_norm": 0.7950716614723206,
|
3325 |
+
"learning_rate": 2.3711340206185566e-06,
|
3326 |
+
"loss": 6.1037,
|
3327 |
+
"step": 948
|
3328 |
+
},
|
3329 |
+
{
|
3330 |
+
"epoch": 9.743589743589745,
|
3331 |
+
"grad_norm": 0.6446979641914368,
|
3332 |
+
"learning_rate": 2.1649484536082473e-06,
|
3333 |
+
"loss": 6.245,
|
3334 |
+
"step": 950
|
3335 |
+
},
|
3336 |
+
{
|
3337 |
+
"epoch": 9.764102564102565,
|
3338 |
+
"grad_norm": 0.7083817720413208,
|
3339 |
+
"learning_rate": 1.9587628865979384e-06,
|
3340 |
+
"loss": 6.1738,
|
3341 |
+
"step": 952
|
3342 |
+
},
|
3343 |
+
{
|
3344 |
+
"epoch": 9.784615384615385,
|
3345 |
+
"grad_norm": 0.611316442489624,
|
3346 |
+
"learning_rate": 1.752577319587629e-06,
|
3347 |
+
"loss": 6.3228,
|
3348 |
+
"step": 954
|
3349 |
+
},
|
3350 |
+
{
|
3351 |
+
"epoch": 9.805128205128206,
|
3352 |
+
"grad_norm": 0.7233178019523621,
|
3353 |
+
"learning_rate": 1.5463917525773197e-06,
|
3354 |
+
"loss": 5.7474,
|
3355 |
+
"step": 956
|
3356 |
+
},
|
3357 |
+
{
|
3358 |
+
"epoch": 9.825641025641026,
|
3359 |
+
"grad_norm": 0.6572598814964294,
|
3360 |
+
"learning_rate": 1.3402061855670102e-06,
|
3361 |
+
"loss": 6.245,
|
3362 |
+
"step": 958
|
3363 |
+
},
|
3364 |
+
{
|
3365 |
+
"epoch": 9.846153846153847,
|
3366 |
+
"grad_norm": 0.7650203108787537,
|
3367 |
+
"learning_rate": 1.134020618556701e-06,
|
3368 |
+
"loss": 6.2761,
|
3369 |
+
"step": 960
|
3370 |
+
},
|
3371 |
+
{
|
3372 |
+
"epoch": 9.866666666666667,
|
3373 |
+
"grad_norm": 0.780173122882843,
|
3374 |
+
"learning_rate": 9.278350515463918e-07,
|
3375 |
+
"loss": 6.2113,
|
3376 |
+
"step": 962
|
3377 |
+
},
|
3378 |
+
{
|
3379 |
+
"epoch": 9.887179487179488,
|
3380 |
+
"grad_norm": 0.7613905668258667,
|
3381 |
+
"learning_rate": 7.216494845360825e-07,
|
3382 |
+
"loss": 6.106,
|
3383 |
+
"step": 964
|
3384 |
+
},
|
3385 |
+
{
|
3386 |
+
"epoch": 9.907692307692308,
|
3387 |
+
"grad_norm": 0.8180409669876099,
|
3388 |
+
"learning_rate": 5.154639175257732e-07,
|
3389 |
+
"loss": 6.1713,
|
3390 |
+
"step": 966
|
3391 |
+
},
|
3392 |
+
{
|
3393 |
+
"epoch": 9.928205128205128,
|
3394 |
+
"grad_norm": 0.6750867366790771,
|
3395 |
+
"learning_rate": 3.0927835051546394e-07,
|
3396 |
+
"loss": 6.1835,
|
3397 |
+
"step": 968
|
3398 |
+
},
|
3399 |
+
{
|
3400 |
+
"epoch": 9.948717948717949,
|
3401 |
+
"grad_norm": 0.7007511854171753,
|
3402 |
+
"learning_rate": 1.0309278350515465e-07,
|
3403 |
+
"loss": 6.2497,
|
3404 |
+
"step": 970
|
3405 |
+
},
|
3406 |
+
{
|
3407 |
+
"epoch": 9.948717948717949,
|
3408 |
+
"step": 970,
|
3409 |
+
"total_flos": 1620553852147860.0,
|
3410 |
+
"train_loss": 6.414023149628,
|
3411 |
+
"train_runtime": 12031.1209,
|
3412 |
+
"train_samples_per_second": 1.296,
|
3413 |
+
"train_steps_per_second": 0.081
|
3414 |
+
}
|
3415 |
+
],
|
3416 |
+
"logging_steps": 2,
|
3417 |
+
"max_steps": 970,
|
3418 |
+
"num_input_tokens_seen": 0,
|
3419 |
+
"num_train_epochs": 10,
|
3420 |
+
"save_steps": 500,
|
3421 |
+
"stateful_callbacks": {
|
3422 |
+
"TrainerControl": {
|
3423 |
+
"args": {
|
3424 |
+
"should_epoch_stop": false,
|
3425 |
+
"should_evaluate": false,
|
3426 |
+
"should_log": false,
|
3427 |
+
"should_save": true,
|
3428 |
+
"should_training_stop": true
|
3429 |
+
},
|
3430 |
+
"attributes": {}
|
3431 |
+
}
|
3432 |
+
},
|
3433 |
+
"total_flos": 1620553852147860.0,
|
3434 |
+
"train_batch_size": 2,
|
3435 |
+
"trial_name": null,
|
3436 |
+
"trial_params": null
|
3437 |
+
}
|